diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index bb0ddacc1339d2a47c1ac9dac07b15a64fa46924..4d857172665091410d9538a67887525705419cbc 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": "20260709T180026Z", - "generated_at": "2026-07-09T18:15:05Z", + "snapshot_id": "20260710T000031Z", + "generated_at": "2026-07-10T00:11:40Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1148,6 +1148,319 @@ } ] }, + { + "cluster_id": "cluster-46376-5", + "summary": "Cluster of 5 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46836, + "issue_numbers": [], + "pr_numbers": [ + 46376, + 46509, + 46836, + 47143, + 47163 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46376, + "right_pr_number": 46509, + "code_similarity": 0.152, + "size_similarity": 0.627, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.176, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 46836, + "code_similarity": 0.071, + "size_similarity": 0.224, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.176, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 47143, + "code_similarity": 0.104, + "size_similarity": 0.431, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.121, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 47163, + "code_similarity": 0.009, + "size_similarity": 0.022, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.028, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 46836, + "code_similarity": 0.119, + "size_similarity": 0.358, + "file_overlap": 0.056, + "area_overlap": 0.0, + "patch_similarity": 0.203, + "shared_filenames": [ + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 412, + 418 + ], + [ + 1071, + 1077 + ], + [ + 1199, + 1205 + ] + ], + "right_ranges": [ + [ + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 1092, + 1098 + ] + ], + "right_ranges": [ + [ + 509, + 515 + ], + [ + 1119, + 1125 + ] + ] + } + ] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47143, + "code_similarity": 0.235, + "size_similarity": 0.687, + "file_overlap": 0.091, + "area_overlap": 0.143, + "patch_similarity": 0.135, + "shared_filenames": [ + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 412, + 418 + ], + [ + 1071, + 1077 + ], + [ + 1199, + 1205 + ] + ], + "right_ranges": [ + [ + 412, + 418 + ], + [ + 975, + 981 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 1092, + 1098 + ] + ], + "right_ranges": [ + [ + 348, + 354 + ], + [ + 1960, + 1966 + ] + ] + } + ] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47163, + "code_similarity": 0.006, + "size_similarity": 0.014, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.024, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46836, + "right_pr_number": 47143, + "code_similarity": 0.156, + "size_similarity": 0.52, + "file_overlap": 0.077, + "area_overlap": 0.0, + "patch_similarity": 0.191, + "shared_filenames": [ + "src/transformers/models/__init__.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/__init__.py", + "left_ranges": [ + [ + 453, + 459 + ] + ], + "right_ranges": [ + [ + 305, + 311 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 + ] + ], + "right_ranges": [ + [ + 412, + 418 + ], + [ + 975, + 981 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 509, + 515 + ], + [ + 1119, + 1125 + ] + ], + "right_ranges": [ + [ + 348, + 354 + ], + [ + 1960, + 1966 + ] + ] + } + ] + }, + { + "left_pr_number": 46836, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.005, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.021, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47143, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.01, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.013, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, { "cluster_id": "cluster-46894-5", "summary": "Cluster of 1 issues and 4 PRs centered on issue #46894.", @@ -1371,494 +1684,162 @@ ], "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 - ] - ] - } - ] - }, - { - "left_pr_number": 46903, - "right_pr_number": 47090, - "code_similarity": 0.519, - "size_similarity": 0.433, - "file_overlap": 0.333, - "area_overlap": 0.867, - "patch_similarity": 0.192, - "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": [ - [ - 637, - 649 - ] - ] - }, - { - "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", - "left_ranges": [ - [ - 500, - 514 - ] - ], - "right_ranges": [ - [ - 500, - 512 - ] - ] - } - ] - }, - { - "left_pr_number": 46958, - "right_pr_number": 47090, - "code_similarity": 0.742, - "size_similarity": 0.852, - "file_overlap": 1.0, - "area_overlap": 0.611, - "patch_similarity": 0.387, - "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": [ - [ - 22, - 28 - ], - [ - 638, - 645 - ] - ], - "right_ranges": [ - [ - 637, - 649 - ] - ] - }, - { - "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", - "left_ranges": [ - [ - 16, - 22 - ], - [ - 501, - 508 - ] - ], - "right_ranges": [ - [ - 500, - 512 - ] - ] - }, - { - "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", - "left_ranges": [ - [ - 15, - 21 - ], - [ - 180, - 221 - ] - ], - "right_ranges": [ - [ - 175, - 222 - ] - ] - } - ] - } - ] - }, - { - "cluster_id": "cluster-44779-5", - "summary": "Cluster of 3 issues and 2 PRs centered on issue #45488.", - "status": "open", - "confidence": 0.6, - "canonical_issue_number": 45488, - "canonical_pr_number": 47017, - "issue_numbers": [ - 44779, - 45488, - 46489 - ], - "pr_numbers": [ - 46580, - 47017 - ], - "evidence_types": [ - "closing_reference", - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46580, - "right_pr_number": 47017, - "code_similarity": 0.241, - "size_similarity": 0.352, - "file_overlap": 0.5, - "area_overlap": 0.0, - "patch_similarity": 0.136, - "shared_filenames": [ - "src/transformers/models/auto/tokenization_auto.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/models/auto/tokenization_auto.py", - "left_ranges": [ - [ - 361, - 367 - ], - [ - 399, - 421 - ], - [ - 724, - 739 - ] - ], - "right_ranges": [ - [ - 428, - 475 - ], - [ - 807, - 825 - ] - ] - } - ] - } - ] - }, - { - "cluster_id": "cluster-46509-4", - "summary": "Cluster of 4 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 46836, - "issue_numbers": [], - "pr_numbers": [ - 46509, - 46836, - 47143, - 47163 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46509, - "right_pr_number": 46836, - "code_similarity": 0.119, - "size_similarity": 0.358, - "file_overlap": 0.056, - "area_overlap": 0.0, - "patch_similarity": 0.203, - "shared_filenames": [ - "src/transformers/models/auto/auto_mappings.py", - "src/transformers/models/auto/modeling_auto.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/models/auto/auto_mappings.py", + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", "left_ranges": [ [ - 412, - 418 - ], - [ - 1071, - 1077 - ], - [ - 1199, - 1205 + 637, + 651 ] ], "right_ranges": [ [ - 637, - 647 - ], - [ - 919, - 928 - ], - [ - 1130, - 1136 + 22, + 28 ], [ - 1260, - 1266 + 638, + 645 ] ] }, { - "filename": "src/transformers/models/auto/modeling_auto.py", + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", "left_ranges": [ [ - 1092, - 1098 + 500, + 514 ] ], "right_ranges": [ [ - 509, - 515 + 16, + 22 ], [ - 1119, - 1125 + 501, + 508 ] ] } ] }, { - "left_pr_number": 46509, - "right_pr_number": 47143, - "code_similarity": 0.235, - "size_similarity": 0.687, - "file_overlap": 0.091, - "area_overlap": 0.143, - "patch_similarity": 0.135, + "left_pr_number": 46903, + "right_pr_number": 47090, + "code_similarity": 0.519, + "size_similarity": 0.433, + "file_overlap": 0.333, + "area_overlap": 0.867, + "patch_similarity": 0.192, "shared_filenames": [ - "src/transformers/models/auto/auto_mappings.py", - "src/transformers/models/auto/modeling_auto.py" + "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/auto/auto_mappings.py", + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", "left_ranges": [ [ - 412, - 418 - ], - [ - 1071, - 1077 - ], - [ - 1199, - 1205 + 637, + 651 ] ], "right_ranges": [ [ - 412, - 418 - ], - [ - 975, - 981 + 637, + 649 ] ] }, { - "filename": "src/transformers/models/auto/modeling_auto.py", + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", "left_ranges": [ [ - 1092, - 1098 + 500, + 514 ] ], "right_ranges": [ [ - 348, - 354 - ], - [ - 1960, - 1966 + 500, + 512 ] ] } ] }, { - "left_pr_number": 46509, - "right_pr_number": 47163, - "code_similarity": 0.006, - "size_similarity": 0.014, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.024, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46836, - "right_pr_number": 47143, - "code_similarity": 0.156, - "size_similarity": 0.52, - "file_overlap": 0.077, - "area_overlap": 0.0, - "patch_similarity": 0.191, + "left_pr_number": 46958, + "right_pr_number": 47090, + "code_similarity": 0.742, + "size_similarity": 0.852, + "file_overlap": 1.0, + "area_overlap": 0.611, + "patch_similarity": 0.387, "shared_filenames": [ - "src/transformers/models/__init__.py", - "src/transformers/models/auto/auto_mappings.py", - "src/transformers/models/auto/modeling_auto.py" + "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/__init__.py", + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", "left_ranges": [ [ - 453, - 459 + 22, + 28 + ], + [ + 638, + 645 ] ], "right_ranges": [ [ - 305, - 311 + 637, + 649 ] ] }, { - "filename": "src/transformers/models/auto/auto_mappings.py", + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", "left_ranges": [ [ - 637, - 647 - ], - [ - 919, - 928 - ], - [ - 1130, - 1136 + 16, + 22 ], [ - 1260, - 1266 + 501, + 508 ] ], "right_ranges": [ [ - 412, - 418 - ], - [ - 975, - 981 + 500, + 512 ] ] }, { - "filename": "src/transformers/models/auto/modeling_auto.py", + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", "left_ranges": [ [ - 509, - 515 + 15, + 21 ], [ - 1119, - 1125 + 180, + 221 ] ], "right_ranges": [ [ - 348, - 354 - ], - [ - 1960, - 1966 + 175, + 222 ] ] } ] - }, - { - "left_pr_number": 46836, - "right_pr_number": 47163, - "code_similarity": 0.004, - "size_similarity": 0.005, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.021, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47143, - "right_pr_number": 47163, - "code_similarity": 0.004, - "size_similarity": 0.01, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.013, - "shared_filenames": [], - "shared_file_areas": [] } ] }, @@ -2007,60 +1988,8 @@ 702 ] ] - } - ] - } - ] - }, - { - "cluster_id": "cluster-46376-3", - "summary": "Cluster of 3 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 47077, - "issue_numbers": [], - "pr_numbers": [ - 46376, - 47077, - 47181 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46376, - "right_pr_number": 47077, - "code_similarity": 0.071, - "size_similarity": 0.299, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.074, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46376, - "right_pr_number": 47181, - "code_similarity": 0.084, - "size_similarity": 0.286, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.179, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47077, - "right_pr_number": 47181, - "code_similarity": 0.029, - "size_similarity": 0.086, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.077, - "shared_filenames": [], - "shared_file_areas": [] + } + ] } ] }, @@ -2168,7 +2097,7 @@ "cluster_id": "cluster-46869-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #46869.", "status": "open", - "confidence": 0.55, + "confidence": 0.75, "canonical_issue_number": 46869, "canonical_pr_number": 46886, "issue_numbers": [ @@ -2179,7 +2108,8 @@ 46922 ], "evidence_types": [ - "closing_reference" + "closing_reference", + "shared_issue_target" ], "pr_comparisons": [ { @@ -2629,6 +2559,59 @@ } ] }, + { + "cluster_id": "cluster-46821-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46821.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 46821, + "canonical_pr_number": 46834, + "issue_numbers": [ + 46821 + ], + "pr_numbers": [ + 46834, + 47223 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 46834, + "right_pr_number": 47223, + "code_similarity": 0.242, + "size_similarity": 0.644, + "file_overlap": 0.333, + "area_overlap": 0.0, + "patch_similarity": 0.088, + "shared_filenames": [ + "src/transformers/models/timesfm2_5/modular_timesfm2_5.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/timesfm2_5/modular_timesfm2_5.py", + "left_ranges": [ + [ + 564, + 570 + ], + [ + 651, + 671 + ] + ], + "right_ranges": [ + [ + 101, + 106 + ] + ] + } + ] + } + ] + }, { "cluster_id": "cluster-46305-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #46305.", @@ -3620,6 +3603,139 @@ } ] }, + { + "cluster_id": "cluster-47122-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47122, + "issue_numbers": [], + "pr_numbers": [ + 47122, + 47150, + 47224 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47122, + "right_pr_number": 47150, + "code_similarity": 0.085, + "size_similarity": 0.423, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47224, + "code_similarity": 0.474, + "size_similarity": 0.302, + "file_overlap": 0.4, + "area_overlap": 0.818, + "patch_similarity": 0.051, + "shared_filenames": [ + "src/transformers/models/edgetam/modeling_edgetam.py", + "src/transformers/models/edgetam/modular_edgetam.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/edgetam/modeling_edgetam.py", + "left_ranges": [ + [ + 464, + 472 + ] + ], + "right_ranges": [ + [ + 464, + 474 + ] + ] + }, + { + "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "left_ranges": [ + [ + 210, + 218 + ] + ], + "right_ranges": [ + [ + 210, + 220 + ] + ] + } + ] + }, + { + "left_pr_number": 47150, + "right_pr_number": 47224, + "code_similarity": 0.442, + "size_similarity": 0.128, + "file_overlap": 0.2, + "area_overlap": 1.0, + "patch_similarity": 0.04, + "shared_filenames": [ + "src/transformers/models/whisper/generation_whisper.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/whisper/generation_whisper.py", + "left_ranges": [ + [ + 1937, + 1943 + ] + ], + "right_ranges": [ + [ + 1937, + 1943 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-47077-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47077, + "issue_numbers": [], + "pr_numbers": [ + 47077, + 47181 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47077, + "right_pr_number": 47181, + "code_similarity": 0.029, + "size_similarity": 0.086, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.077, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, { "cluster_id": "cluster-47034-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", @@ -3649,6 +3765,64 @@ } ] }, + { + "cluster_id": "cluster-44001-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44004, + "issue_numbers": [], + "pr_numbers": [ + 44001, + 44004 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44001, + "right_pr_number": 44004, + "code_similarity": 0.087, + "size_similarity": 0.237, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.262, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-44002-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44002, + "issue_numbers": [], + "pr_numbers": [ + 44002, + 44029 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44002, + "right_pr_number": 44029, + "code_similarity": 0.135, + "size_similarity": 0.43, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.329, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, { "cluster_id": "cluster-44018-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", @@ -3923,77 +4097,9 @@ ] } ] - }, - { - "cluster_id": "cluster-44024-2", - "summary": "Cluster of 2 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 44076, - "issue_numbers": [], - "pr_numbers": [ - 44024, - 44076 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 44024, - "right_pr_number": 44076, - "code_similarity": 0.269, - "size_similarity": 0.976, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.492, - "shared_filenames": [], - "shared_file_areas": [] - } - ] - }, - { - "cluster_id": "cluster-44073-2", - "summary": "Cluster of 2 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 44073, - "issue_numbers": [], - "pr_numbers": [ - 44073, - 44074 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 44073, - "right_pr_number": 44074, - "code_similarity": 0.141, - "size_similarity": 0.58, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.169, - "shared_filenames": [], - "shared_file_areas": [] - } - ] - } - ], - "duplicate_issues": [ - { - "cluster_id": "cluster-44779-5", - "canonical_issue_number": 45488, - "duplicate_issue_numbers": [ - 44779, - 46489 - ], - "reason": "Issues in cluster-44779-5 are treated as duplicates because they share closing_reference, soft_similarity evidence." } ], + "duplicate_issues": [], "duplicate_prs": [ { "cluster_id": "cluster-41211-3", @@ -4040,40 +4146,31 @@ "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, + "cluster_id": "cluster-44001-2", + "canonical_pr_number": 44004, "duplicate_pr_numbers": [ - 44018 + 44001 ], "target_issue_number": null, - "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-44001-2 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-44024-2", - "canonical_pr_number": 44076, + "cluster_id": "cluster-44002-2", + "canonical_pr_number": 44002, "duplicate_pr_numbers": [ - 44024 + 44029 ], "target_issue_number": null, - "reason": "PRs in cluster-44024-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-44002-2 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-44073-2", - "canonical_pr_number": 44073, + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, "duplicate_pr_numbers": [ - 44074 + 44018 ], "target_issue_number": null, - "reason": "PRs in cluster-44073-2 are treated as duplicates because they share soft_similarity evidence." - }, - { - "cluster_id": "cluster-44779-5", - "canonical_pr_number": 47017, - "duplicate_pr_numbers": [ - 46580 - ], - "target_issue_number": 45488, - "reason": "PRs in cluster-44779-5 are treated as duplicates because they converge on issue #45488 with closing_reference, soft_similarity evidence." + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-45081-3", @@ -4121,25 +4218,16 @@ "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-46376-3", - "canonical_pr_number": 47077, - "duplicate_pr_numbers": [ - 46376, - 47181 - ], - "target_issue_number": null, - "reason": "PRs in cluster-46376-3 are treated as duplicates because they share soft_similarity evidence." - }, - { - "cluster_id": "cluster-46509-4", + "cluster_id": "cluster-46376-5", "canonical_pr_number": 46836, "duplicate_pr_numbers": [ + 46376, 46509, 47143, 47163 ], "target_issue_number": null, - "reason": "PRs in cluster-46509-4 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46376-5 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-46693-4", @@ -4151,6 +4239,15 @@ "target_issue_number": 46693, "reason": "PRs in cluster-46693-4 are treated as duplicates because they converge on issue #46693 with closing_reference, shared_issue_target, soft_similarity evidence." }, + { + "cluster_id": "cluster-46821-3", + "canonical_pr_number": 46834, + "duplicate_pr_numbers": [ + 47223 + ], + "target_issue_number": 46821, + "reason": "PRs in cluster-46821-3 are treated as duplicates because they converge on issue #46821 with closing_reference evidence." + }, { "cluster_id": "cluster-46869-3", "canonical_pr_number": 46886, @@ -4158,7 +4255,7 @@ 46922 ], "target_issue_number": 46869, - "reason": "PRs in cluster-46869-3 are treated as duplicates because they converge on issue #46869 with closing_reference evidence." + "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-46894-5", @@ -4180,6 +4277,25 @@ "target_issue_number": null, "reason": "PRs in cluster-47034-2 are treated as duplicates because they share soft_similarity evidence." }, + { + "cluster_id": "cluster-47077-2", + "canonical_pr_number": 47077, + "duplicate_pr_numbers": [ + 47181 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47077-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47122-3", + "canonical_pr_number": 47122, + "duplicate_pr_numbers": [ + 47150, + 47224 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47122-3 are treated as duplicates because they share soft_similarity evidence." + }, { "cluster_id": "cluster-47205-3", "canonical_pr_number": 47218, @@ -4191,10 +4307,10 @@ } ], "best_issue": { - "cluster_id": "cluster-44779-5", - "issue_number": 45488, - "reason": "Issue #45488 is the strongest global issue candidate because it is open, belongs to a cluster with 5 artifacts, and carries 16 discussion comments plus 4 inbound references.", - "score": 186.0 + "cluster_id": "cluster-46894-5", + "issue_number": 46894, + "reason": "Issue #46894 is the strongest global issue candidate because it is open, belongs to a cluster with 5 artifacts, and carries 1 discussion comments plus 9 inbound references.", + "score": 161.0 }, "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 4e3580244d05393d236ec27ee3bef158ea373216..42e43c41f066a512fc4fe4b04fdccba9ee0737a4 100644 --- a/analysis/current/analysis-report-hybrid.llm-reviews.json +++ b/analysis/current/analysis-report-hybrid.llm-reviews.json @@ -1,8 +1,8 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260709T180026Z", - "generated_at": "2026-07-09T18:15:05Z", + "snapshot_id": "20260710T000031Z", + "generated_at": "2026-07-10T00:11:40Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -131,394 +131,382 @@ "issue:45237|issue:46015", "issue:43039|issue:44805" ], - "prepared_review_unit_hash": "8be62a9c0d13fe05e2e9c334a1b0714d8f9ee4e940145e0096fa4e90c1711716", + "prepared_review_unit_hash": "9108cae3644037c8968aa5d06377e9235abba0a236f1b2a1f1b5b35e335f6785", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 62, - "serialized_chars": 46277, - "estimated_input_tokens": 11570, - "estimated_eval_tokens": 23396 + "serialized_chars": 46276, + "estimated_input_tokens": 11569, + "estimated_eval_tokens": 23394 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous, with one clear duplicate subcluster around DeepSpeed ZeRO-3/BERT initialization. Most other pairs are only loosely related by model family, backend, or docs, not the same underlying bug.", - "confidence": 0.72, - "canonical_issue_reason": "issue:43596 is the strongest representative of the only obvious duplicate family: it matches issue:43638 on the same ZeRO-3 'index 0 is out of bounds' failure during BERT initialization.", + "summary": "This cluster is mostly heterogeneous, with only a few true duplicate subgroups: ZeRO-3 load-time index errors, Qwen3.5Moe rotary/offload issues, and a torch 2.4.1 compatibility report. Most other soft pairs are same subsystem but different concrete failures, models, or backends.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43638 is the best representative of the strongest true duplicate family: it describes the shared ZeRO-3 load-time `index 0 is out of bounds for dimension 0 with size 0` failure, which is also reported in issue:43596 and issue:41762.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43596 is the best overall issue anchor here because it is the most specific member of the clearest duplicate pair; the rest of the cluster is too mixed to serve as one unified canonical bug.", + "best_issue_reason": "issue:43638 is the clearest anchor for the cluster\u2019s only well-supported duplicate subgroup, and its title is general enough to represent that specific Zero-3 empty-tensor indexing bug better than the more model-specific variants.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different Qwen3.5 failures: missing config attribute vs streaming chat/completions error." + "reason": "Different failures: config attribute error vs streaming chat/completions API error on a different model." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Unrelated backend issues: AMD GPU failure vs MPS histogram not implemented." + "reason": "Different backend/model problems: AMD GPU failure vs MPS histogram op not implemented for Int." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different problems: image dtype mismatch vs AMD GPU runtime failure." + "reason": "Both are device/backend-related, but the concrete bugs differ: dtype mismatch in multimodal preprocessing vs AMD GPU runtime failure." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Autocast dtype mismatch and AMD GPU failure are different code-path issues." + "reason": "Different concrete dtype-mismatch paths; same broad theme is not enough." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Different models and failures; only broad hardware/runtime overlap." + "reason": "Different models and failure modes; one is an AMD GPU runtime issue, the other a broken example." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "GPT-OSS AMD GPU issue is unrelated to GraniteMoE invalid method call." + "reason": "Different model-specific failures: AMD GPU support vs invalid method call in GraniteMoEHybrid." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch and ModernBERT ONNX export error are unrelated." + "reason": "Different code paths: DeepSpeed collective mismatch during Jamba training vs ONNX export failure for ModernBERT." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "Different models and failure modes: ONNX export vs FineGrainedFP8Config." + "reason": "Different issues in different models; ONNX export bug vs Qwen3Moe FineGrainedFP8 failure." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "ONNX export issue is not the same as Qwen3 auto device-map cuda assert." + "reason": "Different concrete failures: export-time issue vs cudaErrorAssert during auto device mapping." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Table QA crash and RAG docs are not the same bug/change." + "reason": "Docs/example quality issue vs runtime task crash; not the same underlying bug." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "Model export failure vs GenerationMixin import error are distinct." + "reason": "Unrelated failures: export error vs import-time GenerationMixin issue." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions, but they concern different removed features." + "reason": "Both are stale v5 docs, but they reference different removed content and don't look like one concrete duplicate change." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "TRL embedding unpacking and resume/trust-remote-code are different training issues." + "reason": "Different training failures: Qwen3.5 input embedding unpacking vs resume/trust-remote-code behavior." }, { "left": "issue:46165", "right": "issue:46869", "accept": false, - "reason": "Resume-training issue and YOLO multi-GPU device-map failure are unrelated." + "reason": "Different models and symptoms: resume/trust-remote-code issue vs multi-GPU YOLO device-map failure." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Mask shape mismatch is not the same as Gemma4 device-mismatch during forward." + "reason": "Both mention device mismatch, but one is a mask/indexing shape error and the other is a forward device-map mismatch on Gemma4 PLE." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Zero-3 BERT init error and mask/tensor shape mismatch are different bugs." + "reason": "Same general training/runtime area, but the failures are different: ZeRO-3 zero-sized tensor indexing vs mask/tensor shape mismatch." }, { "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "No shared underlying bug beyond both involving model loading/training." + "reason": "Different bugs: ZeRO-3 initialization index error vs resume/trust-remote-code training behavior." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Different failure signatures; one is ZeRO-3 init index error, the other a shape mismatch." + "reason": "Different failure modes; both are errors, but only one is the Zero-3 empty-tensor indexing bug." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Mask indexing error and Qwen3.5 video token/feature mismatch are different." + "reason": "Different multimodal/video chat-template bug vs tensor mask shape mismatch." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Unrelated training-resume issue vs Nemotron example breakage." + "reason": "Different topics: model resume/trust-remote-code vs broken official example." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Mask shape mismatch is not the same as GPT-OSS tensor-parallel failure." + "reason": "Different concrete problems: mask/index shape mismatch vs GPT-OSS tensor-parallel path not working." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video input StopIteration and model-resume/trust-remote-code are unrelated." + "reason": "Different bugs in different areas; video StopIteration vs training resume issue." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice_preset and ModernBERT ONNX export are unrelated." + "reason": "Different model-specific failures: BarkProcessor voice preset vs ModernBERT ONNX export." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Different ZeRO-3 symptoms and different models; not enough to treat as one bug." + "reason": "Both involve distributed training, but the actual bugs differ: collective mismatch vs Zero-3 zero-sized tensor indexing." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 'index 0 is out of bounds' failure during BERT initialization; these look like the same bug." + "reason": "Same Zero-3 load-time `index 0 is out of bounds for dimension 0 with size 0` failure; the model wording differs but the underlying bug and code path match." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy call and GPT-OSS tensor-parallel failure are different issues." + "reason": "Different issues: Liger Kernel cross-entropy routing vs GPT-OSS tensor-parallel failure." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Different problems: detokenization regression vs LoRA merge/save mismatch." + "reason": "Different failures: v5 detokenization regression vs LoRA merge/save/reload inconsistency." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Qwen-3 NVFP4 runtime error and LoRA merge/save inconsistency are unrelated." + "reason": "Different models and bugs; NVFP4 runtime error is unrelated to LoRA merge behavior." }, { "left": "issue:44841", "right": "issue:46661", "accept": false, - "reason": "Processor failure and remote-code config_class AttributeError are different code paths." + "reason": "Processor failure for Voxtral vs remote-code model_class/config_class attribute error." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "DeepSpeed training weight-shape error and collective mismatch are not the same bug." + "reason": "Both are DeepSpeed training issues, but one is a 2-D weight assertion and the other is a collective mismatch on Jamba/ZeRO-3." }, { "left": "issue:43716", "right": "issue:46869", "accept": false, - "reason": "Mistral dtype mismatch and YOLO multi-GPU device-map failure are unrelated." + "reason": "Different models and concrete failures; dtype mismatch in Mistral-3 vs YOLO multi-GPU device-map failure." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpointing/FSDP tensor-size mismatch is not the same as mask-shape indexing error." + "reason": "Both are shape/runtime errors under distributed training, but they are different bugs and code paths." }, { "left": "issue:45685", "right": "issue:46869", "accept": false, - "reason": "MPS MoE histogram error and YOLO device-map auto failure are unrelated." + "reason": "Different backend/model failures: MPS histogram kernel gap vs YOLO multi-GPU device-map error." }, { "left": "issue:45859", "right": "issue:45902", - "accept": false, - "reason": "Same component family, but different concrete failures: CPU offload incompatibility vs uninitialized-memory read after materialization." + "accept": true, + "reason": "Same Qwen3_5Moe rotary embedding offload/materialization bug: one reports CPU offload incompatibility, the other the uninitialized `inv_freq` symptom after `to_empty(cuda)`." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "Activation-checkpointing mismatch and custom trainer loss-signature mismatch are unrelated." + "reason": "Different training bugs: activation-checkpointing/FSDP recomputation mismatch vs `compute_loss` signature change." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "ZeRO-3 BERT init error and GPT-OSS tensor-parallel issue do not share the same bug." + "reason": "Different code paths: ZeRO-3 index error vs GPT-OSS tensor-parallel issue." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "Different model/backend failures; only both mention distributed execution." + "reason": "Different concrete failures; same broad distributed-training label is not enough." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Dtype mismatch in Mistral-3 is unrelated to GraniteMoE invalid method call." + "reason": "Different model-specific failures: dtype mismatch in Mistral-3 vs GraniteMoEHybrid invalid method call." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Qwen-3 NVFP4 runtime failure and AMD GPU failure are different bugs." + "reason": "Different hardware/model bugs; both are runtime failures but not the same underlying issue." }, { "left": "issue:45874", "right": "issue:46294", "accept": false, - "reason": "Gemma4 reverse-embedding lookup and Qwen3.5 video token mismatch are unrelated." + "reason": "Different multimodal/model behaviors: expensive reverse embedding lookup vs video feature/token mismatch." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3 config attribute error with vLLM and Mistral regex patch crash are different issues." + "reason": "Unrelated regressions: Gemma3TextConfig attribute error vs Mistral regex patch crash." }, { "left": "issue:45800", "right": "issue:46515", - "accept": false, - "reason": "Both mention torch 2.4.1, but the reports are too generic to prove the same concrete bug." + "accept": true, + "reason": "Same torch 2.4.1 compatibility report; the second issue is just a shorter restatement of the same incompatibility." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Device mismatch in Gemma4 PLE and expensive reverse embedding lookup are different failures." + "reason": "Different Gemma4 problems: device mismatch during forward vs reverse embedding lookup cost." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch 2.10 incompatibility and Qwen-3 NVFP4 runtime error are unrelated." + "reason": "Different version/model failures; generic torch incompatibility vs 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 not the same underlying bug." + "reason": "Both involve model execution, but one is autocast dtype mismatch and the other is tensor-parallel MoE failure." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Torch dynamo grouped_mm tracing failure and image dtype mismatch are different issues." + "reason": "Different issues: grouped_mm tracing error in Mixtral vs dtype mismatch in Mistral-3 preprocessing." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Different model/runtime failures; same broad numerical area only." + "reason": "Different models and bugs; dtype mismatch under autocast is unrelated to NVFP4 runtime failure." }, { "left": "issue:43716", "right": "issue:45685", "accept": false, - "reason": "Mistral-3 dtype mismatch and MPS histogram-not-implemented are unrelated." + "reason": "Different hardware/backend issues with different model contexts; not the same bug." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid is not the same as the MPS MoE histogram error." + "reason": "Different model/backend failures: invalid method call vs unimplemented MPS histogram op." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Qwen-3 NVFP4 runtime error and TRL SFT embedding unpacking are unrelated." + "reason": "Different failures: NVFP4 runtime issue vs TRL SFT unpacking problem." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different model/runtime errors; both are about model execution but not the same underlying bug." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "Different training paths and failure modes: TRL SFT input unpacking vs GPT-OSS tensor-parallel issue." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger Kernel cross-entropy call and Qwen3.5 embedding unpacking are different issues." + "reason": "Different areas: Liger Kernel cross-entropy call vs Qwen3.5 embedding unpacking in TRL SFT." }, { "left": "issue:41762", "right": "issue:43638", - "accept": false, - "reason": "Both are ZeRO-3 index errors, but on different models and not clearly the same concrete failure." + "accept": true, + "reason": "Same ZeRO-3 `index 0 is out of bounds for dimension 0 with size 0` load-time failure; the model names differ, but the bug pattern matches." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Different docs regressions: pipeline task references vs run command references." + "reason": "Both are stale docs issues, but they cover different removed functionality and look like separate cleanup items." }, { "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 call and autocast dtype mismatch are different bugs." - }, - { - "left": "issue:30277", - "right": "issue:36331", - "accept": false, - "reason": "DeepSpeed collective mismatch and custom trainer signature error are unrelated." - }, - { - "left": "issue:34928", - "right": "issue:36010", - "accept": false, - "reason": "FSDP activation-checkpointing mismatch and GenerationMixin import failure are different issues." - }, - { - "left": "issue:45237", - "right": "issue:46015", - "accept": false, - "reason": "AMD GPU model failure and RAG docs issues are unrelated." + "reason": "Different concrete issues: autocast dtype mismatch vs MPS kernel not implemented." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger Kernel cross-entropy call and tensor mask shape mismatch are different failures." + "reason": "Different bugs: Liger Kernel cross-entropy routing vs mask/tensor indexing shape error." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The canonical issue choice is supported by the strongest duplicate-looking pair (43596/43638), and the soft-edge rejections do not overclaim sameness from weak thematic overlap. No PR-pair mergeability concerns apply here." + "feedback": "The summary is grounded in the packet, and the accepted soft-edge pairs are all based on close title-level matches rather than broad thematic similarity. The conservative rejections are also appropriate. The only mild concern is wording like \"only well-supported duplicate subgroup,\" but the rest of the summary correctly acknowledges multiple small duplicate families, so overall this is acceptable." }, "evaluator_used": true, "retried": false, @@ -592,8 +580,8 @@ "issue:45800|issue:46304", "issue:43296|issue:46304", "issue:43866|issue:46402", - "issue:41762|issue:43596", "issue:39290|issue:41720", + "issue:41762|issue:43596", "issue:43901|issue:44512", "issue:45237|issue:45901", "issue:44589|issue:46304", @@ -632,280 +620,292 @@ "issue:43572|issue:45070", "issue:45800|issue:45823" ], - "prepared_review_unit_hash": "d849ad98a21f82be3c2ef362e4b7a7e2cf3345df4f085c7fe84a29e86f134f7c", + "prepared_review_unit_hash": "c77ae1bc2e9fd62a3e4c89eab6deb7bb32d78061f0707500d64ba55adb25c5a7", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 42, - "serialized_chars": 40037, - "estimated_input_tokens": 10010, - "estimated_eval_tokens": 20276 + "serialized_chars": 40047, + "estimated_input_tokens": 10012, + "estimated_eval_tokens": 20280 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous, but it contains two real duplicate-like threads: a broad v5 tokenizer regression (45812/45920) and a shared ZeRO-3 load-time IndexError pattern (41762/43596). The rest are only topical similarities and should remain separate.", - "confidence": 0.79, - "canonical_issue_reason": "There is no single canonical issue for the whole cluster; most items are unrelated model-specific bugs, docs issues, or environment incompatibilities.", + "summary": "Most items are distinct model- or subsystem-specific bugs. The strongest duplicate signal is the v5 AutoTokenizer regression affecting multiple model families; the rest are too loosely related to merge safely.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45920 is the best representative because it describes the tokenizer regression broadly across multiple model families and clearly frames the underlying v5 token-ID bug.", "canonical_pr_reason": null, - "best_issue_reason": "If one issue must stand in as the representative, 45812 is the best fit because it captures the broad tokenizer regression thread and has the strongest duplicate companion in 45920.", + "best_issue_reason": "issue:45920 is the most actionable and general issue in the set, with the broadest scope for a single fix and the clearest regression characterization.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Different incompatibilities: torch version support vs USE_DISTRIBUTED=0 breakage." + "reason": "Both mention torch/transformers compatibility, but they are different version breakages with different likely causes and no shared concrete code path." }, { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "Different failures: vLLM model loading vs distributed build/runtime support." + "reason": "Different failures: vLLM model loading on transformers 4.53.0 versus PyTorch support broken in 5.9.0. Same ecosystem, not the same bug." }, { "left": "issue:43866", "right": "issue:46402", "accept": false, - "reason": "Different models and causes: corrupted checkpoint vs missing weight loading." - }, - { - "left": "issue:41762", - "right": "issue:43596", - "accept": true, - "reason": "Both are ZeRO-3 load-time IndexErrors with the same empty-tensor/index-0 failure pattern, likely the same shared bug." + "reason": "One is a corrupted checkpoint/model loading problem, the other is missing text encoder weights in Sam3Model. Not the same underlying issue." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Missing config metadata in one case vs CUDA assert during auto device mapping in the other." + "reason": "Both are runtime errors in model execution, but they involve different models, different configs, and different failure modes." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Both involve ZeRO-3 and IndexError, but the model types and reported symptoms differ enough that a shared root cause is not established." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs issues, but they cover unrelated documentation mismatches." + "reason": "Both are docs issues, but they cover different outdated references and do not describe the same change." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "Unrelated subsystems: AMD GPU support vs table QA task crash." + "reason": "Different task and failure class; one is an AMD GPU/model runtime issue, the other is a table-question-answering crash." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to USE_DISTRIBUTED=0 support breakage." + "reason": "Unrelated errors: missing Float8 storage versus distributed-build compatibility. No common bug." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Different code paths: Liger cross-entropy routing vs video StopIteration in chat templating." + "reason": "Liger kernel cross-entropy dispatch and Qwen3-VL video StopIteration are unrelated." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Device-mismatch forward error vs video feature/token mismatch are different bugs." + "reason": "Both are multimodal/model-specific forward-path issues, but one is PLE device mismatch and the other is video token/template mismatch." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Processor failure vs device-map forward mismatch; not the same underlying defect." + "reason": "Processor failure for Voxtral and device-map mismatch for Gemma4 are different code paths." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Detokenization regression vs outdated RAG docs are unrelated." + "reason": "A detokenization regression and outdated RAG examples are unrelated." }, { "left": "issue:43866", "right": "issue:46772", "accept": false, - "reason": "Both are model-related, but the checkpoint/corruption problem is different from NVFP4 compatibility." + "reason": "Corrupted Ovis2 checkpoint versus ModelOpt/NVFP4 compatibility is not the same bug." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "Checkpoint loading bug vs documentation issue; unrelated." + "reason": "NemotronH checkpoint loading and RAG docs cleanup are unrelated." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Remote-code regression is unrelated to the USE_DISTRIBUTED=0 torch support break." + "reason": "Remote-code model breakage and PyTorch USE_DISTRIBUTED compatibility are different underlying problems." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Same underlying tokenizer regression: wrong token IDs across multiple models after v4\u2192v5." + "reason": "Both report the same v5 AutoTokenizer regression causing wrong token IDs across models; they look like the same underlying tokenizer bug and could plausibly be fixed in one PR." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "Model execution failure vs RAG documentation issue." + "reason": "Qwen-3 NVFP4 runtime error and RAG documentation issues are unrelated." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Both are loading failures, but the model-specific causes are different." + "reason": "Checkpoint corruption/load failure versus a model implementation loading issue; no shared fix." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes API incompatibility is unrelated to distributed support breakage." + "reason": "bitsandbytes Int8Params incompatibility and distributed PyTorch support breakage are unrelated." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding_window, but one is a missing config field and the other is a Qwen3-MoE issue." + "reason": "Both mention sliding_window, but the affected models and concrete errors differ; insufficient evidence of the same bug." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are ZeRO-3-related IndexErrors, but the signatures and model setups differ enough to keep separate." + "reason": "Both are ZeRO-3/BERT-related IndexErrors, but one is a mask shape mismatch and the other is an out-of-bounds index; not clearly the same root cause." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper run failure vs table QA crash; unrelated." + "reason": "Whisper loading failure and table-question-answering crash are unrelated." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "Unit-test load failure for one model vs corrupted checkpoint for another." + "reason": "Model test loading failure for GLM-4.7-Flash and a corrupted Ovis2 checkpoint are different issues." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch 2.10 compatibility and pydantic PretrainedConfig regression are different issues." + "reason": "Torch 2.10 incompatibility and a v5.4.0 pydantic config-field regression are not the same bug." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper failure vs pydantic regression; unrelated." + "reason": "Whisper can\u2019t run and pydantic config-field breakage are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization regression versus table QA crash are different problems." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Same model family, but different APIs and failure modes; not one concrete bug." + "reason": "Qwen3.5 chat crash and streaming inference failure are separate code paths with different symptoms." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "Zero-3 load-time IndexError is unrelated to the torch/transformers version incompatibility." + "reason": "Both mention IndexError-style failures, but one is a ZeRO-3 BERT initialization problem and the other is a torch/transformers compatibility issue." }, { "left": "issue:43541", "right": "issue:46869", "accept": false, - "reason": "Grouped_mm tracing error vs multi-GPU device-map failure are different code paths." + "reason": "Both are model execution problems, but Mixtral torch-dynamo grouped_mm tracing and YOLO multi-GPU device mapping are unrelated." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "Qwen NVFP4 runtime failure vs table QA crash are unrelated." + "reason": "Qwen NVFP4 runtime failure and table-question-answering crash do not share an underlying bug." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Different model families and error surfaces." + "reason": "Gemma3 sliding_window config error and Gemma4 device mismatch are distinct issues." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Chat streaming API failure vs MPS histogram kernel limitation are unrelated." + "reason": "Streaming chat API failure and MoE MPS histogram/int type error are unrelated." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "Import/export issue vs config attribute mismatch; not the same bug." + "reason": "ImportError for GenerationMixin and Gemma3 sliding_window config error are different failures." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Grouped_mm tracing failure vs MPS histogram int kernel limitation." + "reason": "Different MoE failures with different operations and symptoms; no shared concrete bug." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both involve MoE/dtype wording, but the execution paths and errors differ." + "reason": "Grouped_mm tracing failure and autocast dtype mismatch are separate runtime issues." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure vs template compilation error; unrelated." + "reason": "Processor failure for Voxtral and template-node compilation error are unrelated." }, { "left": "issue:45507", "right": "issue:46869", "accept": false, - "reason": "Different architectures and failures: invalid method call vs multi-GPU device-map issue." + "reason": "GraniteMoEHybrid invalid method call and YOLO multi-GPU device-map failure are different model-specific bugs." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Weight-tying bug vs corrupted checkpoint." + "reason": "Mistral3 lm_head tying bug and corrupted Ovis2 checkpoint are unrelated." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are IndexError reports, but they affect different models and describe different failure shapes; no clear shared root cause." }, { "left": "issue:45399", "right": "issue:46661", "accept": false, - "reason": "Flash-attn fallback gating and remote-code config_class errors are different subsystems." + "reason": "Flash-attn2 fallback gating and remote-code model_class/config_class handling are different code-path problems." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "StableLm config field regression vs pydantic PretrainedConfig regression." + "reason": "StableLmConfig pad_token_idx regression and pydantic PretrainedConfig breakage are unrelated." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Version incompatibility vs Gemma4 device mismatch are unrelated." + "reason": "Torch/transformers incompatibility and Gemma4 device mismatch are not the same bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately conservative. The two accepted soft edges (45812/45920 and 41762/43596) are supported by closely matching failure surfaces and titles, while the rejected pairs are framed as separate bugs without overclaiming duplicates. No PR-mergeability issue is present here." + "feedback": "Grounded and conservative overall. The canonical choice and summary are supported by the packet, and the one accepted pair (issue:45812 with issue:45920) is clearly the same v5 AutoTokenizer regression. The other verdicts stay appropriately cautious and do not overstate shared root causes." }, "evaluator_used": true, "retried": false, @@ -1007,112 +1007,112 @@ "issue:45183|issue:46463", "issue:44863|issue:45005" ], - "prepared_review_unit_hash": "7a0ffa78d9f6d50dd98e1bc8676b0e08c344238748ace9730ddd95c2bfa57499", + "prepared_review_unit_hash": "676b046d88d2edf0fcc986340146de207c0b38b4b33a01848541b3934ed25d01", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36245, - "estimated_input_tokens": 9062, - "estimated_eval_tokens": 18380 + "serialized_chars": 36243, + "estimated_input_tokens": 9061, + "estimated_eval_tokens": 18378 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a mixed cluster of mostly unrelated bug reports spanning model loading, multimodal/video preprocessing, quantization, and v5 regressions. I only see two soft edges that look like true duplicates: one around Qwen3.5 video-input regression, and one around input-validation error messaging.", - "confidence": 0.71, - "canonical_issue_reason": "Issue 44479 is the best anchor: it describes a concrete v5.3.0 video-input regression affecting several Qwen multimodal variants, and it has a close specific duplicate in 46294.", + "summary": "This cluster is a broad mix of unrelated Transformers/model-serving issues. The pairwise similarities are mostly superficial (shared keywords like dtype, cache, video, or tie weights), so I do not see a trustworthy canonical duplicate or mergeable PR candidate here.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "44479 is the most suitable representative issue because it is broad, actionable, and tied to a clear regression with multiple affected model variants.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different bugs: PaddleOCR-VL loading in vLLM vs adding GGUF support for gpt-oss." + "reason": "Different problems: PaddleOCR-VL load failure in vLLM/Transformers versus GGUF architecture support for gpt-oss." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Both are shape/index failures, but the reported code paths differ: mask/tensor mismatch vs TRL SFT unpacking of embeddings." + "reason": "Both mention shape-related failures, but one is a generic mask/tensor IndexError and the other is a Qwen3.5 + TRL SFT embedding unpacking issue." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Different MoE problems: grouped_mm tracing failure in Mixtral vs tensor-parallel support for GPT-OSS." + "reason": "Different code paths and symptoms: torch dynamo grouped_mm tracing error versus tensor-parallel-only failure on GPT-OSS MoE." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Autocast dtype mismatch and streaming chat/completions failure are distinct symptoms and likely separate paths." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is not the same as Qwen3.5 streaming chat/completions failure." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF architecture support and Qwen3.5 streaming inference are unrelated." + "reason": "GGUF support for gpt-oss and Qwen3.5 streaming inference are unrelated." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "init_empty_weights argument handling and PretrainedConfig/pydantic breakage are different compatibility issues." + "reason": "One is an init_empty_weights argument regression; the other is a PretrainedConfig/pydantic field break." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Memory/OOM regression under int4 quantization is not the same as tied-weights behavior in translation models." + "reason": "Int4 quantization OOM is a different bug from tied-weight handling in translation models." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Sliding-window handling for Qwen3-MoE is unrelated to the pydantic PretrainedConfig regression." + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to the pydantic config-field regression." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are IndexErrors, but one is a mask-length mismatch and the other is a ZeRO-3 init out-of-bounds issue." + "reason": "Both are IndexErrors, but the reported conditions differ: mask length mismatch versus zero-sized tensor during deepspeed zero3 init." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Both involve multimodal inputs, but one is an image-preprocessor dtype mismatch and the other is video token/feature count mismatch." + "reason": "Different modalities and failures: image-preprocessor dtype mismatch versus video feature/token count mismatch." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote-code model breakage and torch/transformers version incompatibility are separate problems." + "reason": "Remote-code compatibility breakage is not the same as a torch 2.4.1 / transformers 5.8.0 incompatibility." }, { "left": "issue:44479", "right": "issue:46294", - "accept": true, - "reason": "These describe the same Qwen3.5 video-input regression: the broad v5.3.0 report and a specific apply_chat_template/video-token mismatch reproduction." + "accept": false, + "reason": "Both involve video input, but the former is a broad v5.3.0 regression across several models while the latter is a specific Qwen3.5 apply_chat_template mismatch." }, { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Same model family, but one is about poorer embeddings and the other about nonsensical AutoModel/pipeline outputs; not enough to prove one bug." + "reason": "Same model family, but one reports poor embedding quality and the other wrong outputs/loading behavior; not clearly the same bug." }, { "left": "issue:46468", @@ -1124,113 +1124,110 @@ "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "ZeRO-3 init indexing and resuming training with trust_remote_code are different failure modes." + "reason": "Zero3 init index error with BertModel is unrelated to resuming training a local remote-code model." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Composite-config unwrapping and uninitialized rotary inv_freq after materialization are different issues." + "reason": "Composite-config unwrap logic and uninitialized-memory rotary embedding are separate issues." }, { "left": "issue:46640", "right": "issue:46661", "accept": false, - "reason": "Flash Attention 2 problem and remote-code config_class AttributeError are distinct." + "reason": "FlashAttention 2 issue and remote-code model_class/config_class error are different failure modes." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Different multimodal/training paths: dtype mismatch in image preprocessing vs TRL SFT embedding unpacking." + "reason": "Image preprocessor dtype mismatch does not match the TRL SFT input-embedding unpacking failure." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP 2-D weight error and TRL input-embedding unpacking are unrelated." + "reason": "FSDP weight dimensionality error and Qwen3.5 TRL unpacking failure are unrelated." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Custom models in notebooks and flash-attn2 fallback checks are separate regressions." + "reason": "Notebook custom-model initialization regression is unrelated to the flash-attn2 fallback gating bug." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "High perplexity on gpt-oss and GLM5 are different model-specific reports." + "reason": "These are different model-specific correctness issues with no shared code-path indicated." }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Both involve custom/remote-code loading, but one is a leaf-directory cache collision and the other is loading the wrong module after save_pretrained; not the same concrete bug." + "reason": "Both touch remote-code/custom-module loading, but one is a cache-path collision and the other is wrong custom module selection after save/load; not clearly the same bug." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "FineGrainedFP8Config failure and flash-attn2 fallback gating are unrelated." + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn2 fallback checks." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Both mention tied weights, but they affect different model families and likely different code paths." + "reason": "Both mention tied weights, but the affected models and concrete failures differ substantially." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "DSA indexer implementation detail vs docs typo are unrelated." + "reason": "A DSA indexer ReLU discussion and a docs typo 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 is a different problem from tied-weight handling." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Cache-strategy typo and removed-run-command documentation are unrelated documentation issues." + "reason": "Cache-strategy typo and removed run-command documentation are unrelated docs issues." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache crash on Qwen3.5 and tied weights in translation models are different issues." + "reason": "PagedAttentionCache linear_attention crash and tied-weight translation-model issues are unrelated." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Input-embedding unpacking in TRL and GraniteMoEHybrid invalid-method calls are unrelated." + "reason": "Qwen3.5 SFT embedding failure and GraniteMoEHybrid invalid-method call are different code paths." }, { "left": "issue:45183", "right": "issue:46463", - "accept": true, - "reason": "Both center on input-validation logic producing wrong/misleading errors; 46463 looks like a T5-specific instance of the broader XOR validation bug." + "accept": false, + "reason": "General input_ids/inputs_embeds XOR validation messaging and T5 decoder-input validation are related only at a high level, not clearly the same bug." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading and tied-weights regressions are unrelated." + "reason": "NemotronH checkpoint loading and tied-weight translation model issues are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "Grounded overall: the cluster is indeed mostly heterogeneous, and the accepted Qwen3.5 video-input pair (44479/46294) is well supported by the titles. The input-validation pair (45183/46463) is also a reasonable conservative duplicate call based on the shared error-message/validation theme. The rejected pairs are generally cautious and do not overstate evidence." - }, - "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 }, @@ -1312,35 +1309,35 @@ "issue:43950|issue:45902", "issue:43054|issue:43493", "issue:44910|issue:45859", - "issue:42947|issue:44387", "issue:45663|issue:45902", + "issue:42947|issue:44387", "issue:44403|issue:44451", "issue:24643|issue:28530", "issue:43975|issue:45520", "issue:44403|issue:44488", - "issue:41553|issue:43883", "issue:43643|issue:45020", "issue:41762|issue:42915", + "issue:41553|issue:43883", "issue:42947|issue:43856", "issue:44991|issue:45399", "issue:43824|issue:43872", "issue:45685|issue:45823", "issue:45081|issue:45399" ], - "prepared_review_unit_hash": "2cbc03a02cd1dc9da6b36ba8563162b26b109fa1960ec8c6dbef8595daf27594", + "prepared_review_unit_hash": "809141961f6eafe3736db428e0476a9f796bc0a817cb2f759a9573481d87a486", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36502, + "serialized_chars": 36504, "estimated_input_tokens": 9126, "estimated_eval_tokens": 18508 }, @@ -1349,201 +1346,201 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous, with only a few true duplicate pairs. The clearest duplicate subclusters are non-persistent buffer corruption, SigLIP2 implementation mismatch, cache.get_seq_length() returning a tensor, and Gemma4 quantized embedder dtype casting.", - "confidence": 0.86, - "canonical_issue_reason": "No single issue covers the whole cluster; 43950 is the best canonical anchor because it is a clear, regression-focused report with a direct duplicate (44534) and a broad enough title to represent its subcluster.", + "summary": "Mostly heterogeneous issue cluster with a few true duplicates. The strongest duplicate family is the non-persistent-buffer regression around #43950/#44534, while #45902 is related thematically but not clearly the same underlying bug from the provided evidence.", + "confidence": 0.64, + "canonical_issue_reason": "#43950 is the clearest anchor for the non-persistent-buffer regression: it names the concrete transformers 5.x corruption bug directly and has an obvious near-duplicate in #44534.", "canonical_pr_reason": null, - "best_issue_reason": "43950 is the strongest global issue target among these items: concrete, clearly bug-shaped, and backed by a near-identical duplicate, while most other items belong to separate bug families.", + "best_issue_reason": "#43950 is the best issue to keep as the representative of the buffer-corruption subset because it is the most explicit and actionable description of that specific regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Different bugs and different model paths; both are tokenizer/runtime failures but not the same underlying issue." + "reason": "Different models and failure modes: tokenizer detokenization regression vs Qwen-3 loading error." }, { "left": "issue:46489", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different failure modes." + "reason": "Both are tokenizer regressions, but they affect different models and do not read as the same concrete bug." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Docs/example errors vs a model quality/discrepancy report; not the same bug." + "reason": "Docs/example errors with quantization versus a model-quality complaint; not the same underlying issue." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Same regression: v5 corrupts non-persistent buffers; 44534 is essentially the same bug described more succinctly." + "reason": "Same regression family: non-persistent buffers being corrupted/junk-filled in transformers v5." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Unrelated failures: custom compute_loss signature mismatch vs DeepSpeed ZeRO-3 load-time index error." + "reason": "Different code paths and symptoms: custom compute_loss API breakage vs DeepSpeed ZeRO-3 loading failure." }, { "left": "issue:43643", "right": "issue:43950", "accept": false, - "reason": "Remote-code config loading bug vs non-persistent buffer corruption; different code paths." + "reason": "Remote-code config field loss is a different bug from non-persistent buffer corruption." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Different backend incompatibilities and different exception classes; not the same change or bug." + "reason": "Distinct compatibility problems with different root causes and symptoms." }, { "left": "issue:46899", "right": "issue:46932", "accept": true, - "reason": "Same Gemma4 quantized embedder dtype-casting bug; 46932 is a follow-up/duplicate of the same issue." + "reason": "Same Gemma4 quantized vision/audio embedder dtype-casting bug; #46932 is explicitly the fix/duplicate." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Different model-specific numerical/dtype issues with different triggers and symptoms." + "reason": "Different issues in different paths: autocast dtype mismatch versus 3D position_ids / SDPA fallback behavior." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding-quality discrepancy vs a broad tokenization-change report; not the same bug." + "reason": "A model embedding quality complaint is not the same as a broad tokenizer-version change report." }, { "left": "issue:45920", "right": "issue:46489", "accept": false, - "reason": "Related tokenizer regressions, but the affected models and concrete failure modes differ enough that they should stay separate." + "reason": "Both involve tokenization, but they are not the same concrete tokenizer bug or model family." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Both mention SigLIP2, but one is zero probabilities and the other is text embedding quality; not clearly the same root cause." + "reason": "Different SigLIP2 symptoms: unexpected zero probabilities versus worse text embeddings." }, { "left": "issue:45987", "right": "issue:46858", "accept": true, - "reason": "Same underlying bug: get_seq_length() returns a tensor instead of a plain int, breaking downstream cache arithmetic." + "reason": "Same StaticCache get_seq_length contract bug: tensor return value breaking downstream arithmetic." }, { "left": "issue:43950", "right": "issue:45902", "accept": false, - "reason": "Both involve buffer state, but one is a general from_pretrained regression and the other is a model-specific to_empty materialization issue." + "reason": "The titles are similar around buffers/materialization, but the packet does not clearly support the same underlying regression; this looks adjacent rather than duplicate." }, { "left": "issue:43054", "right": "issue:43493", - "accept": true, - "reason": "Same SigLIP2 discrepancy family; one reports degraded text embeddings, the other compares HF vs JAX implementations." + "accept": false, + "reason": "Related model family, but one is a performance complaint and the other is an HF-vs-JAX discrepancy report." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Different Qwen3.5 failures: flash-attn packed-sequence misinterpretation vs CPU-offload incompatibility." + "reason": "Separate Qwen3.5 bugs: flash-attn crash from 3D position_ids versus CPU-offload incompatibility in rotary embedding." }, { - "left": "issue:42947", - "right": "issue:44387", + "left": "issue:45663", + "right": "issue:45902", "accept": false, - "reason": "Gradient checkpointing/LoRA behavior vs int4 quantization memory growth; different bugs." + "reason": "Different model/runtime paths and distinct root causes; not enough evidence for a single bug." }, { - "left": "issue:45663", - "right": "issue:45902", + "left": "issue:42947", + "right": "issue:44387", "accept": false, - "reason": "Different model subsystems and failure mechanisms; one is FSDP2 shared_kv_states rebuilding, the other is buffer materialization." + "reason": "Different performance/memory problems with different triggers and code paths." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise vs a specific model load failure; not the same issue." + "reason": "Generic loading noise versus a specific model-loading failure." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "Training with DeepSpeed weight-shape error vs early-stopping metric lookup; unrelated." + "reason": "Unrelated training shape error versus early-stopping metric detection." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Tokenizer detokenization bug vs import_utils flash_attn key error; unrelated." + "reason": "Tokenizer detokenization bug versus import utility KeyError; unrelated." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Both are loading-related, but one is noise/warnings and the other is a concrete model-load failure." - }, - { - "left": "issue:41553", - "right": "issue:43883", - "accept": false, - "reason": "Bad AutoTokenizer error message vs missing all_tied_weights_keys attribute; different code paths." + "reason": "Different loading reports: noisy logs versus a specific tokenizer/model load failure." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "45020 is a broad umbrella about remote_code breakage, but 43643 is a specific config-return-shape bug and not clearly the same root cause." + "reason": "Specific missing-fields issue versus a broader remote-code regression claim." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Different model/config incompatibilities and different exception surfaces." + "reason": "Different model/runtime failures with different root causes." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad error messaging is not the same as a missing attribute error." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Different performance issues: ineffective checkpointing vs MoE memory inefficiency." + "reason": "LoRA gradient checkpointing ineffectiveness and Qwen3 MoE memory usage are different problems." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer load failure vs flash-attn fallback gating; unrelated." + "reason": "Tokenizer loading regression versus flash-attn2 fallback selection logic; different subsystems." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Different incompatibilities: missing Qwen2.5-VL export vs bitsandbytes Int8Params constructor change." + "reason": "Import/export symbol issue versus bitsandbytes constructor incompatibility." }, { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "MPS histogram kernel limitation vs Gemma4 device placement mismatch; unrelated." + "reason": "MPS histogram kernel issue is unrelated to Gemma4 device-map mismatch." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer regex patch crash vs flash-attn fallback checks; not the same bug." + "reason": "Tokenizer regex patch crash versus fallback gating; no shared underlying bug." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and the pairwise calls are generally conservative. The selected canonical anchor (43950) is defensible, and the accepted duplicate/near-duplicate pairs are supported by the issue titles. I don\u2019t see a clear overstatement of evidence." + "accept": false, + "feedback": "The cluster summary is mostly grounded, but one soft-edge acceptance looks overstated: issue:43950 vs issue:45902. The packet only shows title-level similarity around buffer/materialization/uninitialized memory; it does not clearly support calling them the same underlying regression. The other accepted pairs look conservative, but this one makes the verdict set non-conservative overall." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1611,15 +1608,15 @@ ], "soft_pairs": [ "issue:43782|issue:44960", - "issue:41553|issue:43881", "issue:45020|issue:45823", "issue:44661|issue:45020", + "issue:41553|issue:43881", "issue:43828|issue:45823", "issue:45020|issue:45356", "issue:42491|issue:44368", "issue:43881|issue:43883", - "issue:43872|issue:46869", "issue:45715|issue:46315", + "issue:43872|issue:46869", "issue:45759|issue:46661", "issue:43872|issue:44291", "issue:43827|issue:44509", @@ -1634,19 +1631,19 @@ "issue:43940|issue:45313", "issue:43854|issue:44863", "issue:24643|issue:30064", - "issue:44928|issue:45588", "issue:44479|issue:44560", + "issue:44928|issue:45588", "issue:41093|issue:44805", "issue:41553|issue:44991" ], - "prepared_review_unit_hash": "5bf88baecb59abf749bed6d31209f93335c432b193895ef512416129382296d5", + "prepared_review_unit_hash": "3d69cc6b57e3a985e603a9a35e1cd84f903340493141142bb5b0f00694cb9756", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, @@ -1661,177 +1658,189 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose topical mix of unrelated model-loading, tokenizer, DeepSpeed, and audio/video regressions. Only a few soft pairs look like true duplicates: the docs pipeline removal issue, the bitsandbytes `_is_hf_initialized` incompatibility, the DeepSeek/tokenizer regression, the Whisper/fbank failure, and the Qwen VL video regression.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43872 is the best canonical issue here: it is a concrete, well-scoped regression with clear duplicate evidence (44291) and a straightforward compatibility fix path.", + "summary": "Mostly heterogeneous issue cluster with a few clear duplicate pairs. The Qwen video pair is related but not strongly evidenced as the same concrete bug, so it should be treated conservatively rather than merged as a duplicate.", + "confidence": 0.44, + "canonical_issue_reason": "Issue 45020 is the broadest representative of the cluster\u2019s dominant theme: recent-version regressions breaking model/tokenizer loading paths across multiple models. It is not a perfect umbrella, but it is the best single anchor for this mixed set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43872 is the strongest single representative issue in this set because it has a precise failure mode, good discussion signal, and the cleanest duplicate relationship among the accepted pairs.", + "best_issue_reason": "45020 is the strongest general representative because it captures a broad, centrally recurring regression pattern affecting loading behavior across models, which aligns better with the cluster than the narrower model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Different model families and failure modes; both are loading problems, but not the same concrete bug." - }, - { - "left": "issue:41553", - "right": "issue:43881", - "accept": false, - "reason": "Both are loading-related, but one is a tokenizer error message issue and the other is a model loading failure for a different architecture." + "reason": "Different model families and failure modes: Qwen3-VL weight loading with `weight_only=True` vs GLM5 loading in tests." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "One is a remote-code regression, the other is a device-mapping forward mismatch; too different to be the same bug." + "reason": "Both are regressions, but one is remote-code/model loading and the other is a Gemma4 PLE forward device-mismatch bug." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping edge case vs remote-code loading breakage; related area, but not the same underlying issue." + "reason": "Tokenizer mapping/add-new-model failure is a different code path from remote-code loading regressions." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different models and symptoms: Voxtral tokenizer error message vs GLM-4v-9b load failure." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Both mention model execution failures, but the dtype/autocast mismatch and device-map mismatch are distinct code paths." + "reason": "Different runtime bugs in different models: autocast dtype mismatch vs device-map mismatch." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Remote-code loading regressions and tokenizer codec handling are separate problems." + "reason": "Remote-code loading regressions and tokenizer codec/warning regression are different issues." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "LoRA checkpoint compatibility on branch changes is not the same as a warning about `tie_word_embeddings` configuration." + "reason": "LoRA checkpoint compatibility on hf4.x/hf5.x is unrelated to the tie_word_embeddings warning." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Different models and different missing attributes; no evidence of a shared root cause." + "reason": "Both are model-loading failures, but for different models and different missing/tied-weight problems." }, { - "left": "issue:43872", - "right": "issue:46869", + "left": "issue:45715", + "right": "issue:46315", "accept": false, - "reason": "Bitsandbytes init-argument incompatibility is unrelated to a YOLO multi-GPU device-map failure." + "reason": "Both are tokenizer performance regressions, but they affect different conversion functions and are not clearly the same bug." }, { - "left": "issue:45715", - "right": "issue:46315", + "left": "issue:43872", + "right": "issue:46869", "accept": false, - "reason": "Both are tokenizer performance regressions, but one is `convert_ids_to_tokens` and the other is `convert_tokens_to_ids`; not the same bug." + "reason": "Bitsandbytes `_is_hf_initialized` incompatibility is unrelated to YOLO multi-GPU device-map failures." }, { "left": "issue:45759", "right": "issue:46661", "accept": false, - "reason": "Both involve auto-model resolution, but one is `text_config` unwrapping and the other is remote-code `config_class` handling." + "reason": "Different loading bugs: missing `text_config` unwrap for composite configs vs remote-code `config_class` attribute error." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Same compatibility bug: `init_empty_weights` / model loading hits the unexpected `_is_hf_initialized` argument on the same bitsandbytes path." + "reason": "Both describe the same `_is_hf_initialized` compatibility break during model initialization/loading." }, { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "Clear duplicate docs issue: both say removed v5 pipeline tasks are still referenced in documentation." + "reason": "Both are documentation cleanup for stale pipeline-task references removed in v5." }, { "left": "issue:45488", "right": "issue:46489", - "accept": true, - "reason": "Both describe the same v5 tokenizer regression affecting DeepSeek-family tokenization behavior; the later report looks like a concrete manifestation of the same root cause." + "accept": false, + "reason": "Both involve DeepSeek-family tokenizer regressions, but the specific tokenizer behavior and root cause differ." }, { - "left": "issue:43872", - "right": "issue:44291", - "accept": true, - "reason": "Same bitsandbytes/transformers loading incompatibility manifested through the newer `init_empty_weights` path." + "left": "issue:45020", + "right": "issue:46661", + "accept": false, + "reason": "Remote-code model breakage and `config_class` AttributeError are related broadly, but not the same concrete bug." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Tied lm_head weight bug and non-persistent buffer corruption are distinct state/loading problems." + }, + { + "left": "issue:46894", + "right": "issue:46897", + "accept": false, + "reason": "Both are double-shift training-loss reports, but they are model-specific and not clearly the same concrete code-path bug." }, { "left": "issue:43792", "right": "issue:44220", - "accept": true, - "reason": "Both point to the same ASR front-end failure around `_torch_extract_fbank_features()` / Whisper audio feature extraction." + "accept": false, + "reason": "Whisper loading failure and `_torch_extract_fbank_features()` issue are too indirect to treat as the same bug." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Both involve DeepSpeed, but one is about missing weight conversion and the other about gradient averaging under GAS." + "reason": "DeepSpeed ZeRO-3 weight-loading failure and gradient-averaging bug are different training-time issues." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Different runtime errors with unrelated mechanics: trainer loss signature vs mask/tensor shape mismatch." + "reason": "Unexpected keyword in `compute_loss` and mask-shape IndexError are unrelated failures." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "LoRA gradient-checkpointing behavior and a missing capability flag on a remote-code model are not the same bug." + "reason": "One is PEFT LoRA gradient checkpointing not taking effect; the other is a missing `supports_gradient_checkpointing` flag." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are Qwen DeepSpeed load failures, but the model variants and reported failure details are distinct." + "reason": "Both are Qwen DeepSpeed ZeRO-3 load failures, but the affected model structures and missing-weights symptoms are different enough that the shared root cause is not clear." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model-loading failures for unrelated architectures." + "reason": "Different model-loading failures for unrelated model families." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "Completely unrelated: DeepSpeed training weight shape error vs segmentation-map preprocessing." + "reason": "DeepSpeed weight-shape runtime error and void segmentation-map processing are unrelated bugs." }, { - "left": "issue:44928", - "right": "issue:45588", + "left": "issue:44479", + "right": "issue:44560", "accept": false, - "reason": "Gradient explosion/SDPA fallback issue is unrelated to the flash-attention `s_aux=None` crash." + "reason": "Both are Qwen/VL video regressions, but the packet only shows title-level similarity and does not establish the same concrete bug or a single mergeable fix." }, { - "left": "issue:44479", - "right": "issue:44560", - "accept": true, - "reason": "44560 is a specific Qwen3-VL video `StopIteration` symptom under the broader v5.3.0 video-input regression already reported in 44479." + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "RLHF NaNs from position-id behavior and flash-attention `s_aux=None` crash are different problems." }, { "left": "issue:41093", "right": "issue:44805", "accept": false, - "reason": "Same error wording, but no evidence they arise from the same indexed-tensor bug; likely separate reports." + "reason": "The error text is similar, but there is not enough evidence that these are the same underlying mask/indexing bug." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading complaints, but they involve different models and different failure behavior." + "reason": "Different tokenizer-loading problems affecting different models and code paths." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately conservative overall. The accepted soft pairs are plausible duplicates or near-duplicates based on the titles, and the rejected pairs are framed cautiously without overclaiming shared root cause." + "accept": false, + "feedback": "Most of the verdicts are conservative, but one accepted soft edge is too speculative: issue 44479 and issue 44560 are both Qwen/VL video regressions, yet the packet only gives title-level evidence and does not clearly show they are the same concrete bug or a single mergeable PR. The summary also overstates this as a 'duplicate pair' rather than a related pair. Because the soft-edge acceptance is not fully grounded, I would reject." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1852,17 +1861,18 @@ "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", @@ -1905,11 +1915,11 @@ "issue:43366|issue:43531", "issue:43596|issue:46515", "issue:45325|issue:45902", - "issue:41553|issue:44488", "issue:43366|issue:44589", + "issue:41553|issue:44488", "issue:43638|issue:46515", - "issue:45663|issue:45823", "issue:44155|issue:45584", + "issue:45663|issue:45823", "issue:42915|issue:43957", "issue:43824|issue:43957", "issue:43716|issue:44560", @@ -1920,8 +1930,8 @@ "issue:44910|issue:46858", "issue:44560|issue:44918", "issue:36010|issue:42915", - "issue:43550|issue:44910", "issue:44683|issue:46515", + "issue:43550|issue:44910", "issue:45812|issue:46489", "issue:45823|issue:46515", "issue:44661|issue:45823", @@ -1931,247 +1941,247 @@ "issue:43638|issue:44661", "issue:43596|issue:44661", "issue:45672|issue:46899", - "issue:43632|issue:45834", + "issue:42491|issue:43720", "issue:44661|issue:44805" ], - "prepared_review_unit_hash": "bf362bbbf872f91cf3f72a5bc3528204d320ad9a3ebcb7fdfa19a8c3f46c2008", + "prepared_review_unit_hash": "557aeb2856dfa1a9c7f83fde8f2e57f38dd9aae618418bd2461c0c4646bc683c", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 35, - "serialized_chars": 37697, - "estimated_input_tokens": 9425, - "estimated_eval_tokens": 19106 + "serialized_chars": 38341, + "estimated_input_tokens": 9586, + "estimated_eval_tokens": 19428 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous. Only two soft edges look like true duplicates: the custom-model initialization regression (43645 \u2194 43646) and the Siglip2 inference regression (38175 \u2194 43994).", - "confidence": 0.92, - "canonical_issue_reason": "issue:43646 is the broadest clear duplicate target in the set: issue:43645 is just the Jupyter-notebook-specific variant of the same custom model initialization regression.", + "summary": "The set is mostly heterogeneous, with only a few convincing duplicate pairs. The strongest matches are the Transformers 5.0 custom-model init regression, the SigLIP2 bad-output regression, the Deepspeed ZeRO-3 + BertModel IndexError, and the Gemma4 4-bit multimodal dtype-cast failure. Most other soft edges connect different models or different code paths and should stay separate.", + "confidence": 0.87, + "canonical_issue_reason": "issue:46899 is the clearest representative of a concrete, well-scoped regression: Gemma4 multimodal/audio processing breaks under 4-bit quantization due to unsafe dtype handling. It is more specific and actionable than the paired report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best canonical issue among the genuinely duplicated subset because it states the underlying regression more generally; the rest of the cluster is too mixed to share one root cause.", + "best_issue_reason": "issue:46899 is a strong global representative because it describes a precise user-facing failure, has a clear root-cause framing, and aligns with a valid duplicate candidate.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43716", "right": "issue:46468", "accept": false, - "reason": "Different model families and failure modes; not the same bug." + "reason": "Different models and symptoms; one is a Mistral-3 image-preprocessor dtype mismatch, the other is a Nemotron example failure." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "Unrelated issues: GGUF architecture support vs an autocast dtype mismatch." + "reason": "Unrelated bugs: GGUF gpt-oss support vs a Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Different Qwen/Nemotron loading problems with different symptoms and paths." + "reason": "Different failure modes and models: Qwen3.5 PagedAttentionCache crash vs NemotronH checkpoint loading." }, { "left": "issue:44863", "right": "issue:46402", "accept": false, - "reason": "Both are load failures, but for different models and different root causes." + "reason": "Different model-loading issues in unrelated model families." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "One is GGUF architecture support, the other is a sliding_window behavior issue." + "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window handling are separate problems." }, { "left": "issue:43596", "right": "issue:46515", "accept": false, - "reason": "Bert/Deepspeed IndexError vs a generic torch-version incompatibility; not the same bug." + "reason": "No shared bug: ZeRO-3 BertModel IndexError vs a generic torch version incompatibility." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Both involve Qwen internals, but one is VL rope indexing and the other is rotary embedding materialization." + "reason": "Both involve rotary-related code, but the models and root causes differ." }, { - "left": "issue:41553", - "right": "issue:44488", + "left": "issue:43366", + "right": "issue:44589", "accept": false, - "reason": "Different tokenizer/model-loading problems for different models and likely different code paths." + "reason": "Completely different topics: GGUF architecture support vs float8 storage lookup failure." }, { - "left": "issue:43366", - "right": "issue:44589", + "left": "issue:41553", + "right": "issue:44488", "accept": false, - "reason": "GGUF support request vs a float8 storage type error; unrelated." + "reason": "Different loading errors for different models; only broad tokenizer/loading subsystem overlap." }, { "left": "issue:43638", "right": "issue:46515", "accept": false, - "reason": "Specific Deepspeed/Bert IndexError vs broad torch incompatibility; not the same issue." + "reason": "ZeRO-3/BertModel IndexError is unrelated to a generic torch compatibility report." }, { - "left": "issue:45663", - "right": "issue:45823", + "left": "issue:44155", + "right": "issue:45584", "accept": false, - "reason": "Same model family, but distinct bugs: shared_kv_states rebuild vs device-mismatch in PLE." + "reason": "Different audio bugs: track embedding leakage vs empty transcription handling." }, { - "left": "issue:44155", - "right": "issue:45584", + "left": "issue:45663", + "right": "issue:45823", "accept": false, - "reason": "Different audio-generation bugs with different triggers and failure modes." + "reason": "Distinct Gemma4 issues: FSDP2 shared_kv_states rebuild vs device mismatch in PLE forward." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Different regression classes: FP8 config failure vs meta-device loading breakage." + "reason": "Different code paths: Qwen3Moe FP8 config failure vs meta-device loading issues." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Import error for a specific class vs meta-device loading failures; unrelated." + "reason": "Import error for Qwen2.5-VL is unrelated to meta-device loading regressions." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Different Qwen/Mistral-family bugs and different code paths." + "reason": "Different models and errors; no shared underlying bug." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer corruption vs OCR long-input index error; unrelated." + "reason": "Non-persistent buffer loading regression and Kosmos2.5 OCR index error are unrelated." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "3d_parallel runtime issue vs GGUF architecture support; not the same bug." + "reason": "No concrete shared bug; these are different model/runtime issues." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Same custom model initialization regression; 43645 is just the notebook-specific variant." + "reason": "Same Transformers 5.0 custom-model initialization regression; the notebook-specific report is just a narrower reproduction." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Qwen3-vl video StopIteration vs an MPS histogram/int error in MoE code." + "reason": "Unrelated errors in different subsystems and model families." }, { "left": "issue:44910", "right": "issue:46858", "accept": false, - "reason": "Both are cache/position-related, but one is Qwen3.5 flash-attn corruption and the other is static-cache slice arithmetic." + "reason": "Both involve attention/cache internals, but the actual bugs and affected paths are different." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Both mention Qwen3.5/embeddings, but one is video StopIteration and the other is TRL unpacking failure." + "reason": "Different Qwen3.5-related failures with different triggers and code paths." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "Importing GenerationMixin and FP8 config issues are unrelated." + "reason": "Importing GenerationMixin and Qwen3Moe FP8 failures are unrelated." }, { - "left": "issue:43550", - "right": "issue:44910", + "left": "issue:44683", + "right": "issue:46515", "accept": false, - "reason": "Different model-specific attention failures with different stack traces and causes." + "reason": "Compiled flex_attention on torch >=2.9 is not the same bug as a torch 2.4.1 incompatibility report." }, { - "left": "issue:44683", - "right": "issue:46515", + "left": "issue:43550", + "right": "issue:44910", "accept": false, - "reason": "Different torch-version issues: compiled flex_attention on 2.9+ vs a general 2.4.1 incompatibility." + "reason": "Both mention attention/compile, but the model, symptom, and likely root cause differ." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but for different model families and different failure mechanisms." + "reason": "Both are tokenizer regressions, but they affect different model families and likely different code paths." }, { "left": "issue:45823", "right": "issue:46515", "accept": false, - "reason": "Gemma4 device mismatch is unrelated to the torch-version incompatibility report." + "reason": "Gemma4 device mismatch is unrelated to a generic torch incompatibility issue." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "Model-addition/config-mapping failure vs Gemma4 device placement bug; unrelated." + "reason": "add-new-model-like TOKENIZER_MAPPING_NAMES failure and Gemma4 device mismatch are unrelated." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Different shape/indexing failures in different model paths; not a duplicate." + "reason": "Different index/shape errors in different model pipelines." }, { "left": "issue:38175", "right": "issue:43994", "accept": true, - "reason": "Same Siglip2 model regression; 43994 is the broader report covering AutoModel and pipeline." + "reason": "Same SigLIP2-base-patch16-224 bad-output regression; both report incorrect/degenerate outputs from the same model." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Different config-validation failures for different models." + "reason": "Strict config loading for granite_speech and glm4v config validation are separate regressions." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "Deepspeed/Bert IndexError vs tokenizer-mapping model-addition failure; unrelated." - }, - { - "left": "issue:43596", - "right": "issue:44661", - "accept": false, - "reason": "Same as above: different failure mode and subsystem." + "reason": "ZeRO-3 BertModel IndexError is unrelated to tokenizer-mapping/model-addition logic." }, { "left": "issue:45672", "right": "issue:46899", + "accept": true, + "reason": "Both point to the same Gemma4 4-bit multimodal dtype-handling regression; the audio-only and audio/vision descriptions are variant reports of the same bug." + }, + { + "left": "issue:42491", + "right": "issue:43720", "accept": false, - "reason": "Both Gemma4-related, but one is uint8 torch.finfo in audio modules and the other is a broader multimodal embedders quantization bug." + "reason": "Qwen3_moe LoRA compatibility and BitNet accelerate-loading failure are different problems." }, { - "left": "issue:43632", - "right": "issue:45834", + "left": "issue:44661", + "right": "issue:44805", "accept": false, - "reason": "Different regressions: _is_hf_initialized flag vs OCR long-input indexing." + "reason": "No shared underlying bug: tokenizer mapping registration vs mask/index shape mismatch." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analysis is grounded in the packet and stays mostly conservative. The two accepted duplicate pairs are well supported by the titles, and the rest of the soft edges are rejected with appropriately narrow reasons. No obvious overstatement or unsafe merge of unrelated issues." + "feedback": "The accepted duplicate pairs are conservatively phrased and grounded in the packet, and the rejected soft edges are generally cautious rather than overstated. The summary is also supported by the issue titles, including the close Gemma4 and Transformers 5.0 matches. One note: the Deepspeed ZeRO-3/BertModel similarity is mentioned in the summary without a corresponding soft-edge verdict, but that does not appear to overstate the packet evidence." }, "evaluator_used": true, "retried": false, @@ -2200,10 +2210,10 @@ "issue:43257", "issue:43317", "issue:43425", + "issue:43632", "issue:43644", "issue:43645", "issue:43708", - "issue:43720", "issue:43854", "issue:43856", "issue:43873", @@ -2232,6 +2242,7 @@ "issue:45488", "issue:45507", "issue:45812", + "issue:45834", "issue:46050", "issue:46459", "issue:46468", @@ -2241,7 +2252,7 @@ "issue:46901" ], "soft_pairs": [ - "issue:42491|issue:43720", + "issue:43632|issue:45834", "issue:43317|issue:44756", "issue:38175|issue:39692", "issue:43644|issue:44534", @@ -2251,9 +2262,9 @@ "issue:44451|issue:44488", "issue:44960|issue:44977", "issue:44863|issue:46846", - "issue:46459|issue:46512", "issue:45507|issue:46468", "issue:45313|issue:45357", + "issue:46459|issue:46512", "issue:43644|issue:43645", "issue:44488|issue:44991", "issue:39692|issue:43994", @@ -2266,230 +2277,230 @@ "issue:43856|issue:44387", "issue:44960|issue:46050", "issue:44164|issue:44291", - "issue:46897|issue:46901", "issue:45276|issue:45335", + "issue:46897|issue:46901", "issue:45488|issue:45812", "issue:45362|issue:45406", "issue:43425|issue:43957" ], - "prepared_review_unit_hash": "db8e74383f353ba584f355ea62d810ac1e98a3cfbfc7427b05d3b785f88bf229", + "prepared_review_unit_hash": "02ab5b0142f4a884e7e40445dd63f3c22f2cbae122b91652a4fd819f09b57fd9", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 35974, - "estimated_input_tokens": 8994, - "estimated_eval_tokens": 18244 + "serialized_chars": 36464, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues mostly cluster around broad Transformers regressions, but the packet does not provide enough evidence to call any soft pair a definite duplicate. I therefore treat the set conservatively: related in theme, but not the same underlying bug unless the concrete code path is clearer.", - "confidence": 0.88, - "canonical_issue_reason": "issue:43644 is the best canonical representative because it describes a concrete, broad Transformers v5 regression affecting non-persistent buffers, but I am not treating it as a confirmed duplicate anchor here.", + "summary": "This cluster is mostly heterogeneous Transformer/model-loading regressions. Only the non-persistent-buffer junk regression looks like a clear duplicate pair; the other soft pairs mostly share a subsystem or model family but not the same concrete bug.", + "confidence": 0.36, + "canonical_issue_reason": "issue:43644 is the clearest representative of the only strong duplicate pair: it states the broad Transformers v5 non-persistent-buffer corruption bug, while 44534 is the later rephrasing of the same regression.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the strongest global issue candidate because it is broadly scoped and maps to a core runtime regression rather than a model-specific symptom.", + "best_issue_reason": "issue:43644 is the best issue candidate because it is the most general, clearly scoped duplicate target among the accepted pair and has the strongest standalone description.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42491", - "right": "issue:43720", + "left": "issue:43632", + "right": "issue:45834", "accept": false, - "reason": "Both are Qwen3/MoE-related loading failures, but they describe different concrete problems: LoRA compatibility across hf4.x/hf5.x vs packed-weight unpacking during accelerate loading." + "reason": "Unrelated bugs: _is_hf_initialized regression vs Kosmos2.5 OCR index error." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Different failure modes: device_map/offload load failure vs mmap causing OOM on Strix Halo. Same broad memory/loading area, but not the same bug." + "reason": "Different failure modes and code paths: device_map/offload loading vs mmap OOM mitigation." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Same model family, but one is zero probabilities in inference while the other is a documentation example with model/processor mismatch and quantization errors; not the same bug." + "reason": "Both involve SigLIP2, but one is zero probabilities and the other is doc/example mismatch plus quantization failure." }, { "left": "issue:43644", "right": "issue:44534", - "accept": false, - "reason": "The titles suggest a related Transformers v5 buffer-init regression, but the packet does not give enough concrete duplicate evidence to conclude they are the same underlying bug." + "accept": true, + "reason": "Same concrete regression: Transformers v5 fills non-persistent buffers with junk; 44534 is the same bug restated." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Both mention Qwen3, but one is auto device mapping triggering cudaErrorAssert and the other is FineGrainedFP8Config failing on Qwen3Moe. Distinct code paths." + "reason": "Different issues: auto device-mapping CUDA assert vs FineGrainedFP8Config failure." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Unrelated problems: FP8 config failure on Qwen3Moe versus mmap/OOM on Strix Halo." + "reason": "Unrelated Qwen3Moe FP8 config bug vs Strix Halo mmap/OOM workaround." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Both involve Qwen3/MoE-ish loading-era issues only at a very high level; one is LoRA compatibility across branches, the other is resize_token_embeddings not updating Gemma4 embeddings." + "reason": "LoRA hf4.x/hf5.x compatibility problem is unrelated to Gemma4 resize_token_embeddings behavior." }, { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Same broad symptom of model loading failure, but the models and likely root causes differ; cannot treat as the same underlying bug." + "reason": "Both are load failures for different models, but no clear shared underlying bug from the titles alone." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "GLM5 vs Qwen3.5 flash-attention generation failure are different model-specific issues." + "reason": "Different model families and symptoms: GLM5 issue vs Qwen3.5 flash-attention generation bug." }, { "left": "issue:44863", "right": "issue:46846", "accept": false, - "reason": "NemotronH checkpoint loading failure and Qwen3.5 TP-plan missing linear_attn entries are unrelated." - }, - { - "left": "issue:46459", - "right": "issue:46512", - "accept": false, - "reason": "Different API bugs: one ignores dtype when loading composite checkpoints, the other leaves config.dtype stale after from_config(dtype=...)." + "reason": "NemotronH checkpoint loading failure is not the same as Qwen3.5 TP plan OOM/shape errors." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "Different problems in different models: GraniteMoEHybrid invalid method call vs a broken official Nemotron example." + "reason": "GraniteMoEHybrid invalid method call vs Nemotron example breakage are distinct problems." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Both are Qwen3.5 regressions, but one is DeepSpeed ZeRO-3 loading language_model weights and the other is save_pretrained writing incorrect visual encoder keys." + "reason": "Both are Qwen3.5 regressions, but one is ZeRO-3 loading and the other is save_pretrained key serialization." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Different dtype issues: loading composite checkpoints vs from_config leaving config.dtype stale." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Both are Transformers 5 regressions, but one is buffer initialization corruption and the other is Jupyter notebook custom-model initialization; not the same bug." + "reason": "Non-persistent-buffer corruption is unrelated to Jupyter custom-model initialization." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "A model-loading failure for a specific model is not the same as a tokenizer-loading regression for a different model." + "reason": "Different tokenizer/model-loading failures for different model families; no evidence of the same bug." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "SigLIP2 issues, but one is a broken docs example while the other is nonsensical inference results with AutoModel/pipeline; different root causes." + "reason": "Both are SigLIP2-related, but one is doc-example errors and the other is nonsensical inference outputs." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Both are loading/model-integration issues around Qwen3/GLM-style models, but the symptoms are different: deepspeed/accelerate conversion vs unit-test loading failure." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed is unrelated to GLM-4.7-Flash loading in tests." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Different performance/memory topics: inefficient Qwen3 MoE training memory use versus offloading not working as expected with quantization." + "reason": "Training memory inefficiency and quantization/offloading behavior are different bugs." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "LoRA cross-version incompatibility is not the same bug as resize_token_embeddings failing to update output_embeddings." + "reason": "The LoRA compatibility bug is unrelated to generic resize_token_embeddings not updating output embeddings." }, { "left": "issue:42491", "right": "issue:45335", "accept": false, - "reason": "Different model families and different code paths: Qwen3 MoE LoRA compatibility vs T5Gemma embedding resizing." + "reason": "Different model/version-compatibility issue vs t5gemma embedding resize bug." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different models and failure modes: GLM-4.7 test loading issue versus Qwen3-VL weight-shape mismatch." + "reason": "Different models and symptoms: GLM-4.7-Flash load issue vs Qwen3-VL weight-shape mismatch." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Trainer checkpoint resume/max_steps bug and optimizer-argument handling are unrelated." + "reason": "Trainer checkpoint-resume max_steps bug is unrelated to SGD optimizer args being ignored." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Qwen3 MoE training-memory inefficiency is not the same as increased CUDA reserved memory under int4 quantization." + "reason": "Both mention memory, but one is training efficiency and the other is int4 quantization reserved-memory/OOM regression." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "GLM5 model issue and sinq save/load breakage are separate bugs." + "reason": "GLM5 model issue and sinq save/load quantization regression are distinct." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "save/from_pretrained extra_state handling and init_empty_weights unexpected argument are distinct loading/saving bugs." - }, - { - "left": "issue:46897", - "right": "issue:46901", - "accept": false, - "reason": "Both are training-loss shift bugs, but they are reported for different model implementations and do not obviously share the same concrete code path." + "reason": "extra_state save/load handling is different from init_empty_weights/_is_hf_initialized loading TypeError." }, { "left": "issue:45276", "right": "issue:45335", "accept": false, - "reason": "Same general API area, but the affected model implementations differ (Gemma4 vs T5Gemma) and the missing updates are not identical." + "reason": "Similar resize_token_embeddings theme, but different model families and affected submodules; not clearly the same concrete bug." }, { - "left": "issue:45488", + "left": "issue:46897", + "right": "issue:46901", + "accept": false, + "reason": "Both are training-loss shift bugs, but they are model-specific and not proven to be the same code-path issue." + }, + { + "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Tokenizer regressions in v5, but different tokenizer classes, different model families, and different symptoms; too broad to treat as duplicates." + "reason": "Both are tokenizer regressions, but the titles point to different mechanisms and model families, so they are not safely the same bug." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Different crashes: Qwen3.5 chat crash versus Gemma4Processor missing _tokenizer in transformers serve." + "reason": "Qwen3.5 chat crash and Gemma4Processor _tokenizer AttributeError are different failures." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Torch 2.10 incompatibility and meta-device loading breakage are separate environment/runtime issues." + "reason": "Torch 2.10 compatibility and meta-device loading failures are unrelated." } ] }, "evaluator_result": { - "accept": false, - "feedback": "Reject: the analyst overstates evidence on the one positive soft-edge verdict. In the packet, issue:43644 \u2194 issue:44534 has deterministic_accept=false, so it should not be accepted as a clear duplicate pair without stronger support. The summary also claims there is only one clear duplicate pair, but the packet does not provide explicit duplicate evidence for that claim. The rest of the pair judgments are conservative, but this one mismatch is enough to reject." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly duplicative pair in the titles is issue:43644 / issue:44534, which is an exact title match and supports the canonical choice. The other soft-edge rejections are appropriately cautious given the limited evidence in titles/bodies alone." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -2510,6 +2521,7 @@ "issue:39404", "issue:39692", "issue:41720", + "issue:42175", "issue:42796", "issue:42947", "issue:43335", @@ -2517,7 +2529,6 @@ "issue:43404", "issue:43454", "issue:43645", - "issue:43828", "issue:43854", "issue:43873", "issue:43957", @@ -2528,6 +2539,7 @@ "issue:44292", "issue:44387", "issue:44402", + "issue:44451", "issue:44479", "issue:44488", "issue:44589", @@ -2559,8 +2571,8 @@ "issue:44402|issue:44488", "issue:39692|issue:44960", "issue:44291|issue:44756", - "issue:45663|issue:45759", "issue:41720|issue:44155", + "issue:45663|issue:45759", "issue:43404|issue:43454", "issue:42947|issue:45704", "issue:43873|issue:44387", @@ -2574,239 +2586,246 @@ "issue:43335|issue:45663", "issue:44387|issue:44756", "issue:44863|issue:46468", - "issue:44479|issue:44877", "issue:45276|issue:45292", + "issue:44479|issue:44877", "issue:43381|issue:44928", "issue:44164|issue:44756", - "issue:45588|issue:45663", "issue:42796|issue:46326", + "issue:45588|issue:45663", "issue:45684|issue:45698", "issue:43854|issue:45216", - "issue:43828|issue:43957", - "issue:43645|issue:44164", + "issue:42175|issue:43957", "issue:45292|issue:45335", - "issue:44589|issue:44756" + "issue:43645|issue:44164", + "issue:44589|issue:44756", + "issue:44402|issue:44451" ], - "prepared_review_unit_hash": "847f1ff98201286e06a68026876bf61fd5924691d473917b00325fc57539c1a1", + "prepared_review_unit_hash": "469a55762521db6027770b2ed7d879e72925a524e43883e7a369a1d06ad2cc0e", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36567, - "estimated_input_tokens": 9142, - "estimated_eval_tokens": 18540 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37412, + "estimated_input_tokens": 9353, + "estimated_eval_tokens": 18962 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of unrelated Transformers bug reports. The only clear duplicate-like subcluster is around `resize_token_embeddings` not propagating to model-specific embedding/output layers (`issue:45276`, `issue:45292`, `issue:45335`).", - "confidence": 0.84, - "canonical_issue_reason": "`issue:45292` is the broadest and most reusable representative of the resize-token-embedding bug; the other matching reports are model-specific variants of the same failure to update output/decoder embeddings.", + "summary": "Mostly unrelated issue set; the only clear duplicate cluster is the resize_token_embeddings propagation bug across related model variants.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 45292 is the cleanest canonical: it states the shared resize_token_embeddings/output_embeddings bug in the most general form, while the neighboring reports are model-specific variants.", "canonical_pr_reason": null, - "best_issue_reason": "`issue:45292` best captures the root problem in a model-agnostic way, making it the strongest cluster representative for triage and linkage to the related model-specific reports.", + "best_issue_reason": "Issue 45292 is the best representative because it is the broadest and most reusable description of the underlying bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both involve flash-attention crashes, but the failure modes differ: 3D `position_ids` packed-sequence misinterpretation vs `s_aux=None` attribute error on sink-less models." + "reason": "Both are Flash Attention bugs, but they fail in different code paths and for different causes." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Both are model-loading/tokenizer complaints, but one is a vocab-size mismatch and the other is a separate load failure for a different model." + "reason": "Both are model-loading failures, but the reports name different models and do not show the same root cause." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Different models and different problems: SigLIP2 doc-example errors vs a generic GLM5 issue." + "reason": "Different models and failure modes; one is a doc/example issue, the other is a GLM5 report." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "`init_empty_weights` argument handling and Strix Halo mmap OOM are unrelated bugs." + "reason": "One is init_empty_weights argument handling; the other is mmap/OOM on Strix Halo." }, { - "left": "issue:45663", - "right": "issue:45759", + "left": "issue:41720", + "right": "issue:44155", "accept": false, - "reason": "Different code paths: Gemma4 FSDP2 attention state handling vs Qwen multimodal `text_config` unwrapping." + "reason": "Completely different subsystems and symptoms: device-map CUDA assert vs batched audio inference leakage." }, { - "left": "issue:41720", - "right": "issue:44155", + "left": "issue:45663", + "right": "issue:45759", "accept": false, - "reason": "One is a Qwen3 auto device-map CUDA assert; the other is an AudioFlamingo3 batched inference leak. Same broad area, different bugs." + "reason": "Different model families and bugs: FSDP2 keyerror vs from_config text_config unwrapping." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Similar symptom class (untied `lm_head`), but different model families and not enough evidence of the same underlying code-path bug." + "reason": "Both are weight-tying bugs, but for different model implementations and likely different fixes." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/LoRA behavior and an apex RMSNorm memory leak are unrelated." + "reason": "Gradient checkpointing/LoRA behavior and an Apex RMSNorm memory leak are unrelated." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both mention quantization and memory pressure, but one is offloading behavior and the other is reserved-memory/OOM growth." + "reason": "Both mention quantization memory issues, but one is offloading and the other is int4 reserved-memory growth." }, { "left": "issue:44079", "right": "issue:46829", "accept": false, - "reason": "Different output-structure bugs in different model classes; the shared wording is too generic to treat as the same issue." + "reason": "ModelOutput key assignment and VideoPrism output formatting are different bugs." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Gemma4 FSDP2 `shared_kv_states` rebuild and Qwen3_5Moe rotary-embedding CPU offload are unrelated." + "reason": "Different model code paths: Gemma4 FSDP2 vs Qwen3_5Moe CPU offload incompatibility." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "`init_empty_weights` loading error and MXFP4 quantization fallback are separate problems." + "reason": "Both are quantization/loading problems, but they concern different models and quantization formats." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Both involve gradient checkpointing, but one is PEFT/LoRA effectiveness and the other is eval-mode restrictions." + "reason": "Same broad topic of training behavior, but one is checkpointing inefficacy and the other is eval-mode restriction." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "Import path breakage for `GenerationMixin` is unrelated to Whisper pipeline `return_language` behavior." + "reason": "ImportError for GenerationMixin and Whisper pipeline return_language are unrelated." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Meta-device loading issues and Strix Halo mmap OOM are distinct." + "reason": "Meta-device loading regressions and mmap/OOM on Strix Halo are different problems." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "AudioFlamingo3 token leak and Gemma4 PLE device mismatch are unrelated." + "reason": "AudioFlamingo batching leak and Gemma4 device mismatch are unrelated." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "Different model families and different failures; only a broad training/configuration similarity exists." + "reason": "SwitchTransformers sparse-layer creation and Gemma4 FSDP2 shared_kv_states are unrelated." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Quantization OOM under int4 and mmap OOM on Strix Halo are not the same bug." + "reason": "Quantization OOM and mmap-on-Strix-Halo OOM have different causes and code paths." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "Both are Nemotron-related, but one is checkpoint loading and the other is an example breakage." - }, - { - "left": "issue:44479", - "right": "issue:44877", - "accept": false, - "reason": "Video-input regression in Qwen VL models and strict config loading for granite_speech are unrelated." + "reason": "NemotronH checkpoint loading and the official Nemotron example break in different ways." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same root bug: `resize_token_embeddings` does not propagate to model-specific output/embedding layers; `45276` is a Gemma4-specific instance of the generic issue in `45292`." + "reason": "Both report resize_token_embeddings failing to propagate new embeddings to output embeddings; same underlying bug." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video input regression in Qwen variants and strict config loading for granite_speech are unrelated." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Gradient-checkpointing eval-mode behavior is unrelated to the Qwen3.5 RLHF NaN/position-id issue." + "reason": "Eval-mode checkpointing restriction and Qwen3.5 RLHF gradient explosions are different issues." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "Save/load `extra_state` handling and Strix Halo mmap OOM are unrelated." + "reason": "save/from_pretrained extra_state handling and Strix Halo mmap OOM are unrelated." }, { - "left": "issue:45588", - "right": "issue:45663", + "left": "issue:42796", + "right": "issue:46326", "accept": false, - "reason": "Flash-attention sinkless-model crash and Gemma4 FSDP2 key error are different code paths." + "reason": "Both involve mistral_common tokenizers, but one is BPE marker handling and the other is continue_final_message logic." }, { - "left": "issue:42796", - "right": "issue:46326", + "left": "issue:45588", + "right": "issue:45663", "accept": false, - "reason": "Both concern tokenizer behavior, but BPE-marker handling and `continue_final_message` are separate tokenizer bugs." + "reason": "Flash-attn sink-less crash and Gemma4 FSDP2 keyerror are different code paths." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Both involve save/load of custom models, but one is permissions propagation and the other is wrong-module loading after save." + "reason": "One is file permission propagation, the other is loading the wrong custom module after save_pretrained." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different models and different regression surfaces: unit-test loading failure vs checkpoint-save regression." + "reason": "GLM-4.7-Flash test loading and Qwen3.5 save_pretrained regression are different bugs." }, { - "left": "issue:43828", + "left": "issue:42175", "right": "issue:43957", "accept": false, - "reason": "Autocast dtype mismatch and meta-device loading failures are not the same issue." - }, - { - "left": "issue:43645", - "right": "issue:44164", - "accept": false, - "reason": "Jupyter custom-model initialization and `save/from_pretrained` extra_state handling are separate problems." + "reason": "Package dependency/back-end installation and meta-device loading regressions are unrelated." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same underlying `resize_token_embeddings` propagation bug; `45335` is another model-specific manifestation where decoder embeddings are not updated." + "reason": "Both are resize_token_embeddings propagation failures in related model variants; same underlying class of bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Jupyter custom-model initialization and extra_state save/load handling are unrelated." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to the Strix Halo mmap/OOM issue." + "reason": "Missing Float8 storage and Strix Halo mmap OOM are different failure modes." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both are tokenizer/model loading complaints, but not clearly the same root cause or code path." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: most items are clearly unrelated by title/body metadata, and the only strongly supported subcluster is the resize_token_embeddings propagation bug around 45276/45292/45335. The conservative false verdicts are appropriate, and the two true soft-edge merges are supported by nearly identical issue descriptions." + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted cluster, the resize_token_embeddings propagation bug across 45292/45276/45335, is well supported by the titles and looks like a single mergeable bugfix family. The other false verdicts do not overclaim shared root cause from the available evidence." }, "evaluator_used": true, "retried": false, @@ -2829,6 +2848,7 @@ "nodes": [ "issue:36032", "issue:39404", + "issue:43299", "issue:43425", "issue:43452", "issue:43596", @@ -2841,12 +2861,11 @@ "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", @@ -2878,14 +2897,13 @@ ], "soft_pairs": [ "issue:43646|issue:43708", - "issue:44402|issue:44451", "issue:46894|issue:46901", "issue:43425|issue:43828", "issue:46315|issue:46491", "issue:45812|issue:46315", "issue:44530|issue:45084", - "issue:43425|issue:46869", "issue:39404|issue:45701", + "issue:43425|issue:46869", "issue:43611|issue:43645", "issue:44683|issue:45823", "issue:43425|issue:43716", @@ -2895,8 +2913,8 @@ "issue:44479|issue:44805", "issue:43901|issue:44509", "issue:43994|issue:44079", - "issue:44164|issue:45759", "issue:43906|issue:44530", + "issue:44164|issue:45759", "issue:36032|issue:43452", "issue:44265|issue:44898", "issue:46050|issue:46326", @@ -2908,228 +2926,229 @@ "issue:45479|issue:46894", "issue:44164|issue:46821", "issue:43856|issue:44756", - "issue:43596|issue:44683" + "issue:43596|issue:44683", + "issue:43299|issue:43931" ], - "prepared_review_unit_hash": "e3181b5ddd21009611bdd32bfe67eed0513281a4fa01abc4f8161638a2b2f249", + "prepared_review_unit_hash": "d4c5dbc9da89caf066374905557800f3e3dbacc8033d7af3b7c036011b851cd8", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37546, - "estimated_input_tokens": 9387, - "estimated_eval_tokens": 19030 + "serialized_chars": 37541, + "estimated_input_tokens": 9386, + "estimated_eval_tokens": 19028 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a grab bag of unrelated Transformers regressions across different code paths (tokenizers, model loading, multimodal input, training, export, docs). I do not see any soft pair that is safe to merge as the same underlying bug/change, so all soft edges are rejected. The best representative issue is the broad custom-model initialization regression.", - "confidence": 0.73, - "canonical_issue_reason": "issue:43646 is the broadest, most representative user-facing regression in the set: a general Transformers 5.0.0 custom-model initialization failure, which is closer to a family of adjacent v5 custom-model breakages than the more narrow, model-specific reports.", + "summary": "The cluster is dominated by heterogeneous Transformers v5 regressions (tokenizers, model loading, multimodal dtype/device issues, and a few training/runtime bugs). None of the soft-edge pairs look like the same underlying defect closely enough to merge, so all proposed duplicate links should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "issue:45812 is the strongest canonical issue: it describes a concrete, broad v4\u2192v5 regression in tokenizer behavior with clear user impact across all Granite models, making it a better representative target than the more model-specific or more generic-loading issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best anchor because it describes a general breakage rather than a model-specific symptom, making it the most plausible canonical issue for the surrounding v5 custom-model-loading/init regression family.", + "best_issue_reason": "issue:45812 is the best overall issue in the set because it is concrete, broadly scoped, and clearly tied to a specific regression in a core code path (AutoTokenizer), which makes it a useful anchor for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Different bug class: custom model initialization in v5 vs Trainer resume/max_steps math." - }, - { - "left": "issue:44402", - "right": "issue:44451", - "accept": false, - "reason": "Both are tokenizer-loading related, but one is a vocab/token setup mismatch and the other is a model-specific load failure; no clear shared root cause." + "reason": "Different failure modes: custom model init in notebooks vs Trainer resume/max_steps calculation. Same broad v5 breakage theme, but not the same bug." }, { "left": "issue:46894", "right": "issue:46901", "accept": false, - "reason": "Same symptom pattern, but they affect different model implementations and do not clearly point to the same code path." + "reason": "Both mention a double-shift pattern in training loss, but they affect different model families and no shared code-path is shown; too weak to treat as one bug." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "One is a Torch version incompatibility report; the other is an autocast dtype-mismatch bug in a specific model." + "reason": "One is a Torch 2.10 compatibility complaint, the other is an autocast dtype mismatch in Phi-tiny-MoE. Different root causes and fixes." }, { "left": "issue:46315", "right": "issue:46491", "accept": false, - "reason": "Performance regression in token conversion vs a tokenization round-trip correctness bug." + "reason": "Performance regression in convert_tokens_to_ids vs a round-trip tokenization correctness bug for CodeLlama. Not the same issue." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Wrong token IDs for Granite models is a correctness regression, not the slow-path performance issue in convert_tokens_to_ids." + "reason": "Tokenizer IDs are wrong for Granite models, while this pair is about conversion performance for slow tokenizers. Different bug classes." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "Different failures: PagedAttentionCache linear_attention crash vs a template-compilation TypeError." + "reason": "PagedAttentionCache linear_attention crash vs a template-node compilation error. Both are runtime failures, but on different code paths and subsystems." }, { - "left": "issue:43425", - "right": "issue:46869", + "left": "issue:39404", + "right": "issue:45701", "accept": false, - "reason": "Torch 2.10 incompatibility is unrelated to a YOLO multi-GPU device_map failure." + "reason": "Whisper pipeline return_language regression vs a general tokenization-change report. Too broad and not the same underlying defect." }, { - "left": "issue:39404", - "right": "issue:45701", + "left": "issue:43425", + "right": "issue:46869", "accept": false, - "reason": "Whisper pipeline return_language regression is not the same as broad tokenization behavior changes." + "reason": "Torch version incompatibility and YOLO multi-GPU device_map failure are unrelated problems." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Both mention custom models, but one is about base_model_prefix loading and the other about notebook initialization behavior; different concrete failures." + "reason": "base_model_prefix loading error and notebook custom-model initialization are both v5 model-loading issues, but they are not the same concrete bug." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention failure on newer Torch vs Gemma4 device-mismatch in forward are separate code paths." + "reason": "Compiled flex_attention on torch >=2.9 and Gemma4 device mismatch with device_map=\"auto\" are separate runtime issues." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "General Torch compatibility report vs a multimodal dtype mismatch in Mistral-3." + "reason": "Torch incompatibility vs Mistral-3 image preprocessor/model dtype mismatch. Different causes and code paths." }, { "left": "issue:39404", "right": "issue:46665", "accept": false, - "reason": "Whisper pipeline behavior regression and from_pretrained-in-class inconsistency are unrelated." + "reason": "Whisper pipeline regression and from_pretrained behavior inside Python classes are unrelated loading/pipeline issues." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "No shared concrete bug: Torch incompatibility vs Gemma4 device_map forward mismatch." + "reason": "A Torch version incompatibility is not the same as a model forward device-placement bug." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Storage-object lookup failure and non-template-node compilation failure are different issues." + "reason": "Float8 storage lookup failure and template-node compilation failure are distinct low-level errors." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video-input regression in several Qwen VL variants is not the same as a generic mask shape mismatch error." + "reason": "Video-input regression for Qwen VL models vs a mask shape mismatch. Both are regressions, but not the same defect." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs issues, but they concern different removed/misdocumented pipeline APIs and are not duplicates." + "reason": "Both are docs-related, but one is about stale return_all_scores text and the other about removed pipeline task docs. Not the same issue." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Nonsensical Siglip2 outputs and ModelOutput key assignment are unrelated bugs." + "reason": "Nonsensical SigLIP2 outputs and ModelOutput key assignment with None are different behavioral bugs." }, { - "left": "issue:44164", - "right": "issue:45759", + "left": "issue:43906", + "right": "issue:44530", "accept": false, - "reason": "Save/from_pretrained extra_state handling and composite Qwen config unwrapping are different serialization/loading problems." + "reason": "Isolated reproduction of another issue vs a PagedAttentionCache crash. The first is just a reproduction report, not the same code-path bug." }, { - "left": "issue:43906", - "right": "issue:44530", + "left": "issue:44164", + "right": "issue:45759", "accept": false, - "reason": "An isolated reproduction of an old issue is not the same as a Qwen3.5 cache crash." + "reason": "extra_state serialization failure and composite Qwen config unwrapping are unrelated save/load behaviors." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "T5 tokenizer special_tokens conflict and gguf_file from_pretrained breakage are different loading failures." + "reason": "T5 tokenizer add_special_tokens conflict and gguf_file failures in from_pretrained are both loading problems, but they arise from different mechanisms." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export export failure due to torch_compilable_check is unrelated to Perceiver interpolation behavior." + "reason": "torch.export failure due to torch_compilable_check vs Perceiver non-default-resolution interpolation failure. Different model families and failure modes." }, { "left": "issue:46050", "right": "issue:46326", "accept": false, - "reason": "Quantization save/load breakage vs continue_final_message behavior on mistral_common tokenizers are different bugs." + "reason": "Quantized sinq save/load corruption and mistral_common continue_final_message behavior are unrelated." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "Old-issue reproduction and mistral_common continue_final_message regression do not share a concrete code-path problem." + "reason": "A reproduction of another issue and a tokenizer prompt-construction bug are not the same underlying defect." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Tokenizer load regression for one model and a backend_tokenizer attribute error in a Mistral regex patch are different tokenizer bugs." + "reason": "Tokenizer loading failure for EMBEDDIA/est-roberta vs a Mistral regex patch crash. Both are tokenizer regressions, but on different code paths." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT MLM prediction regression is not the same issue as EMBEDDIA/est-roberta tokenizer loading failure." + "reason": "CamemBERT prediction corruption and tokenizer load failure for est-roberta are distinct issues." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper pipeline behavior and MXFP4 quantization fallback are unrelated." + "reason": "Whisper return_language pipeline regression vs MXFP4 quantization fallback issue. Different models and logic." }, { "left": "issue:45823", "right": "issue:46899", "accept": false, - "reason": "Both are Gemma4-related, but one is a device placement mismatch and the other is a 4-bit dtype-cast bug in multimodal embedders." + "reason": "Gemma4 device mismatch during forward vs Gemma4 multimodal quantization dtype-cast bug. Related model family, but distinct failures and fixes." }, { "left": "issue:45479", "right": "issue:46894", "accept": false, - "reason": "Sequence-classification zero-loss pathology and ASR training-loss double-shift are different model families and likely different implementations." + "reason": "Zero-loss classification bug and ASR double-shift bug are different training-loss problems with no clear shared code path." }, { "left": "issue:44164", "right": "issue:46821", "accept": false, - "reason": "extra_state save/load handling and TimesFM window_size AttributeError are unrelated." + "reason": "extra_state save/load bug vs TimesFM window_size AttributeError. Different APIs and root causes." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Training memory inefficiency and Strix Halo mmap OOM avoidance are not the same bug." + "reason": "Qwen3 MoE memory inefficiency and Strix Halo mmap OOM are both memory-related, but not the same defect." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "Deepspeed ZeRO3/BertModel index error and torch>=2.9 flex_attention compilation failure are unrelated." + "reason": "Zero3/BertModel index error and compiled flex_attention torch>=2.9 failure are unrelated runtime problems." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3VL MoE loading breakage and Qwen3-VL-30B-A3B-Instruct weight-shape mismatch are similar only at a family level, but they are not the same concrete bug or fix." } ] }, @@ -3162,7 +3181,7 @@ "issue:42175", "issue:42796", "issue:42915", - "issue:43299", + "issue:43317", "issue:43404", "issue:43421", "issue:43452", @@ -3176,7 +3195,6 @@ "issue:43645", "issue:43646", "issue:43716", - "issue:43931", "issue:44206", "issue:44220", "issue:44451", @@ -3186,12 +3204,14 @@ "issue:44661", "issue:44683", "issue:44737", + "issue:44811", "issue:44987", "issue:44991", "issue:45081", "issue:45084", "issue:45479", "issue:45507", + "issue:45584", "issue:45859", "issue:46178", "issue:46285", @@ -3202,9 +3222,8 @@ "issue:47050" ], "soft_pairs": [ - "issue:43632|issue:47050", "issue:44488|issue:44987", - "issue:43299|issue:43931", + "issue:43632|issue:47050", "issue:45479|issue:46897", "issue:44488|issue:45081", "issue:43611|issue:43644", @@ -3216,230 +3235,238 @@ "issue:44661|issue:44683", "issue:32090|issue:39404", "issue:44451|issue:44987", - "issue:43541|issue:45084", "issue:43643|issue:44530", + "issue:43541|issue:45084", "issue:42915|issue:44530", "issue:44737|issue:45859", "issue:40990|issue:43421", "issue:34567|issue:40444", "issue:43644|issue:43646", "issue:42175|issue:46178", - "issue:43582|issue:43632", "issue:40444|issue:46326", "issue:36032|issue:43645", + "issue:43582|issue:43632", "issue:44206|issue:44220", "issue:44987|issue:44991", "issue:43452|issue:43482", - "issue:45084|issue:45507" + "issue:44811|issue:45584", + "issue:45084|issue:45507", + "issue:43317|issue:43482" ], - "prepared_review_unit_hash": "abf7f12b19a456b02d4a6d268eacbd2f42a67ec85198120c487eb050f11b0fc2", + "prepared_review_unit_hash": "b665da7647bb5d814690c8d4a5616f4407550c9e0bb12e1978cb6ca6d2804fe3", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35922, - "estimated_input_tokens": 8981, - "estimated_eval_tokens": 18218 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36791, + "estimated_input_tokens": 9198, + "estimated_eval_tokens": 18652 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many issues are unrelated Transformers regressions across different models, tokenizers, and runtime paths. Only a couple of soft pairs look like the same underlying bug in a shared code path.", - "confidence": 0.73, - "canonical_issue_reason": "issue:43646 is the broadest representative of the cluster\u2019s recurring Transformers v5 model-initialization regressions, and is the best anchor for the group.", + "summary": "Most items are distinct regressions across tokenizer loading, model initialization, audio feature extraction, and compile/runtime paths. I found two plausible duplicate pairs that appear to describe the same concrete loader/helper bug rather than just the same subsystem.", + "confidence": 0.67, + "canonical_issue_reason": "Issue 43644 is a clear, self-contained Transformers v5 regression with a concrete symptom and broad impact, making it a reasonable anchor issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the most generally applicable, user-facing issue in the set; it can stand in for several nearby v5 initialization failures better than the narrower symptom reports.", + "best_issue_reason": "Issue 43644 is the strongest representative issue here: it describes a specific v5 regression, is easy to understand, and has some discussion/reference activity.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43632", - "right": "issue:47050", - "accept": false, - "reason": "Different failures: a v5 model-loading/init flag regression vs a Canine ONNX attention dtype mismatch." - }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading failures, but they affect different models and likely different root causes." + "reason": "Both are tokenizer/model loading failures, but they affect different repositories and likely different root causes." }, { - "left": "issue:43299", - "right": "issue:43931", + "left": "issue:43632", + "right": "issue:47050", "accept": false, - "reason": "Same family, but one is a generic Qwen3VL-MoE load break and the other is a specific weight-shape mismatch on a different variant." + "reason": "These are unrelated bugs: an initialization flag regression versus an ONNX attention dtype mismatch." }, { "left": "issue:45479", "right": "issue:46897", "accept": false, - "reason": "Both mention training-loss issues, but they are different models and different bug patterns." + "reason": "Both mention training loss problems, but they involve different models and the titles do not indicate the same code path." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Tokenizer/model loading in different subsystems; no clear shared code path." + "reason": "Different tokenizer-loading bugs: one is a repo loading failure, the other is a Mistral regex patch crash." }, { "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." + "reason": "Both are v5 regressions, but one is about base_model_prefix loading and the other about non-persistent buffer initialization." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "Unrelated: T5 tokenizer method conflict vs Qwen3Moe FP8 loading failure." + "reason": "These are unrelated tokenizer and FP8-config failures." }, { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Both concern tokenizers, but one is repository tokenizer file selection and the other is a specific DeepSeek-Coder v1 output regression." + "reason": "Both concern tokenizers, but one is about ignoring tokenizer.json while the other is an output regression in a specific model tokenizer." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Different Mistral-family bugs: tokenizer BPE handling vs lm_head weight tying." + "reason": "Different Mistral-related bugs: tokenizer BPE markers versus tied lm_head weights in a conditional generation model." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Completely different areas: image preprocessor dtype mismatch vs template compilation failure." + "reason": "These are different failures with no clear shared code path: image preprocessor dtype mismatch versus template compilation errors." }, { "left": "issue:46285", "right": "issue:46402", "accept": false, - "reason": "Different model classes and failures: CLIP structure inconsistency vs Sam3 text encoder weights not loading." + "reason": "Different model-structure/weight-loading regressions." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "No relation between tokenizer-mapping behavior and torch flex_attention compilation." + "reason": "An add-new-model registry bug and a flex_attention compile regression are unrelated." }, { "left": "issue:32090", "right": "issue:39404", "accept": false, - "reason": "Trainer GPU broadcast error is unrelated to Whisper pipeline return_language behavior." + "reason": "Trainer GPU broadcast and Whisper return_language pipeline issues are unrelated." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading issues, but they involve different models and no shared root cause is evident." + "reason": "Both are loading failures, but they appear to be different models and different tokenizer/model paths." }, { - "left": "issue:43541", - "right": "issue:45084", + "left": "issue:43643", + "right": "issue:44530", "accept": false, - "reason": "Both mention runtime/compiler failures, but the code paths are unrelated (grouped_mm tracing vs non-template node compilation)." + "reason": "A trust_remote_code AutoConfig regression is unrelated to the Qwen3.5 PagedAttentionCache crash." }, { - "left": "issue:43643", - "right": "issue:44530", + "left": "issue:43541", + "right": "issue:45084", "accept": false, - "reason": "Different bugs: missing fields from trust_remote_code config loading vs PagedAttentionCache group-type crash." + "reason": "Both involve compilation/tracing, but the concrete failures and code paths are different." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Both touch Qwen-family models, but one is FP8 config loading and the other is a cache/group-type runtime crash." + "reason": "Different Qwen-family bugs affecting different features and models." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Different tensor/device bugs in unrelated models: XLNet CPU arange vs Qwen3_5Moe CPU offload incompatibility." + "reason": "Both are device/offload-related, but the model-specific bugs are distinct." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity regression on a model is unrelated to tokenizer post-processor update behavior." + "reason": "Perplexity/calibration issues and runtime post-processor updates are not the same bug." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "Trainer token-count tracking and multi-image iterable dataset failures are unrelated." + "reason": "Unrelated trainer token-count tracking and iterable-dataset multi-image failure." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Both are v5 init regressions, but one is junk buffers and the other is custom model initialization; too broad to merge safely." + "reason": "Both are v5 initialization regressions, but the titles point to different symptoms and no shared concrete code path." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "Both mention TensorFlow backend setup, but one is install extras missing tensorflow and the other is a BACKENDS_MAPPING error for tensorflow_text." - }, - { - "left": "issue:43582", - "right": "issue:43632", - "accept": false, - "reason": "Apple Silicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." + "reason": "These are different backend-mapping/package-installation issues." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Different problems: multi-image finetuning failure vs mistral_common continue_final_message behavior." + "reason": "A multimodal training failure and a mistral_common tokenizer message-handling bug are unrelated." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Tokenizer special-token method conflict is unrelated to custom model initialization in notebooks." + "reason": "T5 tokenizer method conflicts and custom-model initialization in notebooks are different problems." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "An Apple Silicon allocator TypeError is unrelated to the `_is_hf_initialized` regression." }, { "left": "issue:44206", "right": "issue:44220", "accept": true, - "reason": "Both point to the same audio feature-extraction path; the specific LasrFeatureExtractor crash appears to be a symptom of the underlying `_torch_extract_fbank_features()` issue." + "reason": "These look like the same audio feature-extraction regression: one names the unsupported center arg, the other points to the internal `_torch_extract_fbank_features()` helper that would surface it." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are versioned loading regressions, but they affect different models and likely different tokenizer/config paths." + "reason": "Both are v5 tokenizer-loading failures, but they involve different models and no shared concrete failing path is evident." }, { "left": "issue:43452", "right": "issue:43482", "accept": true, - "reason": "Both are GGUF loading failures under transformers v5, plausibly the same broken `gguf_file` loading path affecting tokenizer/model from_pretrained." + "reason": "Both are GGUF loading regressions under Transformers v5; 43482 looks like a specific model manifestation of the broader gguf_file/from_pretrained breakage in 43452." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both involve Whisper special-token behavior, but one is batch_decode skipping tokens and the other is empty-transcription generation after align_special_tokens." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "No shared concrete code path: template compilation failure vs GraniteMoEHybrid invalid method call." + "reason": "A template compilation error and a GraniteMoEHybrid invalid method call are different bugs." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "GPU/CPU offload loading failure and GGUF loading failure are different regression types." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and appropriately conservative. The accepted soft pairs are plausible from the titles alone, and the rejected pairs avoid overstating shared root causes. The choice of issue:43646 as the canonical broad representative is defensible given the surrounding v5 initialization regressions." + "feedback": "The summary is grounded in the packet and stays mostly conservative. The two accepted soft pairs are plausible duplicates based on the titles and share a concrete bug theme without overclaiming a shared code path. I don\u2019t see any clear evidence inflation that would justify rejecting." }, "evaluator_used": true, "retried": false, @@ -3463,14 +3490,16 @@ "issue:28530", "issue:34567", "issue:36032", + "issue:41720", "issue:42915", "issue:43208", "issue:43262", - "issue:43317", + "issue:43322", "issue:43377", "issue:43404", - "issue:43482", + "issue:43425", "issue:43531", + "issue:43540", "issue:43632", "issue:43645", "issue:43646", @@ -3488,18 +3517,17 @@ "issue:44451", "issue:44464", "issue:44466", + "issue:44479", "issue:44488", "issue:44534", "issue:44589", "issue:44610", - "issue:44811", "issue:44849", "issue:44877", "issue:44987", "issue:45081", "issue:45084", "issue:45325", - "issue:45584", "issue:45987", "issue:46326", "issue:46661", @@ -3509,8 +3537,6 @@ "issue:47013" ], "soft_pairs": [ - "issue:44811|issue:45584", - "issue:43317|issue:43482", "issue:43746|issue:46326", "issue:43262|issue:43377", "issue:43792|issue:46869", @@ -3522,211 +3548,227 @@ "issue:43377|issue:43819", "issue:44849|issue:45325", "issue:44987|issue:45081", + "issue:28530|issue:34567", "issue:42915|issue:44291", "issue:43404|issue:44466", - "issue:28530|issue:34567", - "issue:43883|issue:44488", "issue:36032|issue:43742", "issue:44451|issue:45081", + "issue:43883|issue:44488", "issue:43994|issue:44610", "issue:44062|issue:44987", "issue:44464|issue:45987", "issue:42915|issue:44589", "issue:44008|issue:46899", + "issue:44265|issue:44610", "issue:43531|issue:44877", + "issue:36032|issue:41720", "issue:43632|issue:46821", - "issue:44265|issue:44610" + "issue:43322|issue:43540", + "issue:44479|issue:45325", + "issue:43425|issue:44403" ], - "prepared_review_unit_hash": "87a281d70b0eaeda702df1b120c89477fc66965ae2babc117f1439bf69e15d05", + "prepared_review_unit_hash": "81ce28a1efea6e47d2d9ed4e2b322abd696c1f8a707a0620d19b3b8f8c03c617", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35445, - "estimated_input_tokens": 8862, - "estimated_eval_tokens": 17980 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36564, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is broadly heterogeneous: it spans Whisper, Qwen, xLSTM, tokenizer loading, model init, quantization, and cache/export issues. None of the soft pairs look like the same underlying bug or change, so I reject all proposed duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "issue:45081 is the clearest single bug report in the set: it has a specific crash, clear repro context, and multiple inbound references. That said, the overall cluster is not a true duplicate group.", + "summary": "This cluster is a loose collection of unrelated closed issues with superficial text similarity (mostly Transformers v5, multimodal models, tokenizer loading, and xLSTM). None of the soft pairs look like the same underlying bug/change, so no duplicates should be merged.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue: the items span distinct bugs in different code paths and model families, so collapsing them would be misleading.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45081 is the strongest representative issue here because it is concrete, well-scoped, and externally referenced; the other items are too diverse to serve as a shared canonical.", + "best_issue_reason": "No issue is a good global representative of this cluster; the similarity is mostly topical rather than duplicate-level.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44811", - "right": "issue:45584", - "accept": false, - "reason": "Both are Whisper-related, but one is about batch_decode ignoring skip_special_tokens and the other is an empty-transcription failure after align_special_tokens; different bugs." - }, - { - "left": "issue:43317", - "right": "issue:43482", - "accept": false, - "reason": "Both are loading regressions, but one concerns dequantized model offload with device_map=auto and the other GGUF loading under v5; different code paths and causes." - }, { "left": "issue:43746", "right": "issue:46326", "accept": false, - "reason": "PEFT local-checkpoint loading for GraniteSpeech is unrelated to continue_final_message behavior in mistral_common tokenizers." + "reason": "Both involve model loading/tokenization workflows, but one is PEFT local checkpoint loading for GraniteSpeech while the other is mistral_common continue_final_message behavior; different bugs." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio-processing themed, but one is a wrong default sampling rate and the other is a missing padding-mask causing batched/single mismatch." + "reason": "Audio processor chat-template sampling-rate defaulting and MIMI batched-vs-single output mismatch are unrelated failures in different components." }, { "left": "issue:43792", "right": "issue:46869", "accept": false, - "reason": "Whisper inference failure and YOLO multi-GPU device_map failure are unrelated models and failure modes." + "reason": "Whisper load failure and YOLO multi-GPU device-map failure are separate model/runtime issues with different symptoms and code paths." }, { "left": "issue:43208", "right": "issue:47013", "accept": false, - "reason": "The left issue is a broad xLSTM umbrella for three bugs; the right is one specific train-mode unpack crash. Not clearly the same underlying bug." + "reason": "Both are xLSTM, but one reports multiple training bugs for small models and the other is a specific return_last_states unpacking crash; not the same change." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "One is generic loading noise/logging, the other is a tokenizer backend AttributeError in Mistral regex patching." + "reason": "Generic loading noise and a mistral regex patch crash are different problems; one is user-facing logging, the other tokenizer patch logic." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and non-template-node compilation error are different problems." + "reason": "Qwen3Moe FineGrainedFP8Config failure and non-template-node compilation error are distinct multimodal/compiler issues." }, { "left": "issue:43645", "right": "issue:46661", "accept": false, - "reason": "Notebook custom-model initialization breakage is not the same as remote-code AutoModel missing config_class." + "reason": "Both concern custom/remote-code model initialization, but one is notebook/custom model init with v5, while the other is missing config_class on remote-code models; different root causes." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization failures and non-persistent buffer corruption are distinct v5 regressions." + "reason": "Custom model initialization breakage and non-persistent buffer junk are separate Transformer v5 regressions." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "MIMI padding-mask mismatch and DAC.from_latents vs forward mismatch are different models and different bugs." + "reason": "MIMI padding-mask batching mismatch is unrelated to DAC.from_latents vs forward/STE consistency." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Different Qwen models and different symptoms: hidden_states output bug versus temporal position_ids scaling bug." + "reason": "Qwen3.5 hidden-states bug and Qwen2.5-VL rope-index scaling bug affect different models and behaviors." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Both are loading failures in v5-era code, but one is a specific model load regression and the other is a tokenizer regex crash." + "reason": "Loading physical-intelligence/fast failing on v5.1.0 is unrelated to the mistral tokenizer regex patch crash." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Early stopping/eval metric handling and TrainerState token counting are different Trainer bugs with no shared code path indicated." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "FineGrainedFP8Config failure is unrelated to init_empty_weights passing an unexpected _is_hf_initialized argument." + "reason": "Both mention v5 / initialization, but FineGrainedFP8Config Qwen3Moe failure is unrelated to init_empty_weights and _is_hf_initialized argument handling." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "One is missing tied lm_head weights in Mistral3, the other is inconsistent tied-weight serialization by device." - }, - { - "left": "issue:28530", - "right": "issue:34567", - "accept": false, - "reason": "Early-stopping metric selection and TrainerState token-count tracking are unrelated training bugs." - }, - { - "left": "issue:43883", - "right": "issue:44488", - "accept": false, - "reason": "A missing all_tied_weights_keys attribute and inability to load a specific BERT model are not the same issue." + "reason": "Mistral3 lm_head weight tying and device-dependent serialization of lm_head.weight are related only at a broad 'tied weights' level, but they describe different bugs and fixes." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "T5 tokenizer special-token conflict and MobileLLM key error are different loader/tokenizer failures." + "reason": "T5 tokenizer add_special_tokens conflict and MobileLLM key error are separate tokenizer/model-loading issues." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT loading failure and Mistral tokenizer backend crash are unrelated model-specific regressions." + "reason": "ScandiBERT loading failure and mistral regex patch crash involve different models and failure modes." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Molmo missing all_tied_weights_keys and sleng-bert loading failure are unrelated model-loading regressions." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Siglip2 nonsensical outputs and OmDet-Turbo processor size mismatch affect different pipelines and root causes." + "reason": "Siglip2 nonsensical outputs and OmDet-Turbo processor size mismatch are different vision-model bugs." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "AddedToken special-argument collision is a tokenizer construction bug, not the same as the physical-intelligence/fast load failure." + "reason": "AddedToken special-argument duplication and physical-intelligence/fast loading failure are unrelated tokenizer vs model-loading issues." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Compiled chunked generation inconsistency and StaticCache return-type mismatch are separate cache/generation issues." + "reason": "Chunked generation inconsistency under compiled forward and StaticCache.get_seq_length contract violation are different generation/cache bugs." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "Qwen3Moe FP8 config failure and missing Float8 storage object are different quantization/runtime errors." + "reason": "Qwen3Moe FP8 config failure and Float8 storage lookup failure are distinct quantization/runtime issues." }, { "left": "issue:44008", "right": "issue:46899", "accept": false, - "reason": "Gemma3n variable-name collision in forward and Gemma4 unsafe dtype casting are distinct multimodal bugs." + "reason": "Gemma 3n audio/vision variable collision and Gemma4 multimodal 4-bit dtype cast issue are different multimodal bugs." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export with torch_compilable_check and OmDet-Turbo processor resolution mismatch are unrelated export vs preprocessing problems." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and strict granite_speech config validation are unrelated." + "reason": "Qwen3-MoE sliding_window bug and granite_speech strict config loading are separate model/config issues." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "T5 tokenizer add_special_tokens conflict and Qwen3 auto device mapping cudaErrorAssert are unrelated." }, { "left": "issue:43632", "right": "issue:46821", "accept": false, - "reason": "The _is_hf_initialized flag breakage and TimesFM window_size AttributeError are different v5 regressions." + "reason": "_is_hf_initialized flag breakage and TimesFM window_size AttributeError are different Transformers v5 regressions in separate subsystems." }, { - "left": "issue:44265", - "right": "issue:44610", + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Llava Next segmentation fault and Qwen3OmniMoe video-input ValueError are distinct multimodal loading/processing bugs." + }, + { + "left": "issue:44479", + "right": "issue:45325", "accept": false, - "reason": "torch.export/exportability with torch_compilable_check is unrelated to OmDet-Turbo producing the wrong image resolution." + "reason": "Qwen2.5/VL video-input regression and Qwen2.5-VL rope-index temporal scaling bug are both VL-related but affect different code paths and symptoms." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch 2.10 incompatibility is a dependency-version issue; unnecessary loading noise is a logging/polish issue." } ] }, @@ -3752,21 +3794,18 @@ "nodes": [ "issue:33666", "issue:35707", - "issue:36032", "issue:40444", "issue:40990", "issue:41720", + "issue:42175", "issue:42796", "issue:43089", "issue:43296", - "issue:43322", "issue:43381", "issue:43421", - "issue:43425", "issue:43482", "issue:43525", "issue:43531", - "issue:43540", "issue:43611", "issue:43632", "issue:43716", @@ -3777,7 +3816,7 @@ "issue:44220", "issue:44291", "issue:44403", - "issue:44479", + "issue:44462", "issue:44534", "issue:44568", "issue:44589", @@ -3789,6 +3828,8 @@ "issue:45216", "issue:45325", "issue:45357", + "issue:45538", + "issue:45542", "issue:45663", "issue:45812", "issue:45923", @@ -3796,29 +3837,26 @@ "issue:46315", "issue:46491", "issue:46710", + "issue:46821", "issue:46899" ], "soft_pairs": [ - "issue:43322|issue:43540", - "issue:36032|issue:41720", - "issue:44479|issue:45325", - "issue:43425|issue:44403", "issue:43089|issue:45987", "issue:43296|issue:43531", "issue:45216|issue:45357", - "issue:43611|issue:44534", "issue:40444|issue:45325", + "issue:43611|issue:44534", "issue:44220|issue:44683", "issue:44991|issue:46315", "issue:43716|issue:46899", - "issue:43883|issue:44991", "issue:44403|issue:44991", + "issue:43883|issue:44991", "issue:43381|issue:45923", "issue:42796|issue:43746", "issue:45812|issue:46491", "issue:43792|issue:44589", - "issue:43632|issue:44756", "issue:44291|issue:44589", + "issue:43632|issue:44756", "issue:33666|issue:40444", "issue:40990|issue:44811", "issue:42796|issue:46710", @@ -3827,216 +3865,234 @@ "issue:44079|issue:45663", "issue:43531|issue:44291", "issue:43421|issue:44568", - "issue:41720|issue:45081" + "issue:41720|issue:45081", + "issue:42175|issue:45542", + "issue:43482|issue:45538", + "issue:44462|issue:45812", + "issue:43296|issue:46821", + "issue:44079|issue:44534", + "issue:42175|issue:43296" ], - "prepared_review_unit_hash": "65970d998e076611926770f7bc77ed52fd896fa9741d61422bebd4c135d1da53", + "prepared_review_unit_hash": "1e057b8b5e33ac06e8fcbc0c981d7f85423903e44290fdbe60123e3e617789de", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35922, - "estimated_input_tokens": 8981, - "estimated_eval_tokens": 18218 + "soft_pair_count": 31, + "serialized_chars": 36498, + "estimated_input_tokens": 9125, + "estimated_eval_tokens": 18506 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous and does not form one duplicate cluster. The only likely duplicate family is the Qwen3.5 save_pretrained regression pair; everything else looks like separate bugs or regressions across different models/code paths.", - "confidence": 0.79, - "canonical_issue_reason": "Issue 45216 is the broadest and clearest representative of the only plausible duplicate family: a Qwen3.5 save_pretrained regression. Issue 45357 reads like a narrower follow-up focused on visual encoder keys.", + "summary": "This set is mostly heterogeneous rather than one tight duplicate cluster. Only a couple of pairs look like true duplicates: the Qwen3.5 save_pretrained regression pair and the TensorFlow-backend packaging pair.", + "confidence": 0.76, + "canonical_issue_reason": "No single canonical issue fits the whole cluster; the items span unrelated tokenizer, generation, multimodal, packaging, and model-loading bugs. If forced to pick one umbrella issue for the only clear subcluster, issue 45216 is the best representative because it is the broader Qwen3.5 save_pretrained regression and 45357 looks like a narrower follow-up.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45216 is the best global representative because it states the regression broadly, is clearly scoped to a concrete API failure, and most likely subsumes the narrower 45357 report.", + "best_issue_reason": "Issue 45216 is the strongest issue candidate: it is a clear, user-facing regression report with a broader scope and it plausibly subsumes 45357. The rest of the items are too mixed to share one canonical issue.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43322", - "right": "issue:43540", - "accept": false, - "reason": "Different models and failure modes: Llava Next loading segfault vs Qwen3OmniMoe video ValueError." - }, - { - "left": "issue:36032", - "right": "issue:41720", - "accept": false, - "reason": "Unrelated tokenizer constructor conflict vs Qwen3 auto device mapping CUDA assert." - }, - { - "left": "issue:44479", - "right": "issue:45325", - "accept": false, - "reason": "Both are Qwen VL regressions, but one is a broad video-input regression and the other is a still-image rope-index scaling bug." - }, - { - "left": "issue:43425", - "right": "issue:44403", - "accept": false, - "reason": "Torch version incompatibility is unrelated to noisy model-loading logs." - }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation performance overhead is unrelated to the StaticCache return-type bug." + "reason": "Different problems: generation runtime overhead vs a StaticCache return-type contract bug." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "Different models and code paths: PaddleOCR-VL loading in vLLM vs Qwen3-MoE sliding-window behavior." + "reason": "Different model families and failure modes: PaddleOCR-VL loading in vLLM vs Qwen3-MoE sliding_window behavior." }, { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Both describe the same Qwen3.5 save_pretrained serialization regression; 45357 appears to be a narrower symptom around visual encoder keys." + "reason": "Both report the same Qwen3.5 save_pretrained regression; 45357 looks like a narrower symptom of incorrect checkpoint contents." }, { - "left": "issue:43611", - "right": "issue:44534", + "left": "issue:40444", + "right": "issue:45325", "accept": false, - "reason": "Different v5 breakages: base_model_prefix loading vs non-persistent buffer initialization." + "reason": "Both are Qwen2.5-VL-related, but one is multi-image training failure and the other is rope/position_id scaling for still images." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:43611", + "right": "issue:44534", "accept": false, - "reason": "Multi-image iterable dataset finetuning failure is not the same bug as still-image position-id scaling." + "reason": "Both involve Transformers v5 regressions, but they affect different code paths: model loading with base_model_prefix vs non-persistent buffer initialization." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio feature extraction bug vs compiled flex_attention incompatibility are unrelated." + "reason": "Unrelated subsystems: fbank feature extraction vs compiled flex_attention on newer torch." }, { "left": "issue:44991", "right": "issue:46315", "accept": false, - "reason": "Tokenizer loading failure and slow-tokenizer conversion performance regression are distinct issues." + "reason": "Both touch tokenizers in v5, but one is tokenizer loading correctness and the other is slow-tokenizer performance with many added tokens." }, { "left": "issue:43716", "right": "issue:46899", "accept": false, - "reason": "Both involve dtype issues in multimodal models, but they affect different models and different failure paths." + "reason": "Both are multimodal dtype issues, but they involve different models and different failure paths (preprocessor/model mismatch vs unsafe cast in 4-bit embedders)." }, { - "left": "issue:43883", + "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Missing all_tied_weights_keys on Molmo is unrelated to tokenizer loading for est-roberta." + "reason": "Different bugs: noisy model loading vs tokenizer loading failure for a specific repo." }, { - "left": "issue:44403", + "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Loading noise is not the same underlying tokenizer-loading failure." + "reason": "Different model-loading attribute error vs tokenizer-loading regression." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Eval-mode gradient checkpointing misuse is different from a missing supports_gradient_checkpointing flag." + "reason": "One is about checkpointing being disallowed in eval mode; the other is a missing supports_gradient_checkpointing flag." }, { "left": "issue:42796", "right": "issue:43746", "accept": false, - "reason": "Tokenizer BPE marker handling and PEFT checkpoint loading are unrelated." + "reason": "Different issues: tokenizer BPE marker handling vs PEFT/local-checkpoint loading for GraniteSpeech." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but one changes token IDs broadly for Granite while the other is a CodeLlama round-trip spacing issue." + "reason": "Both are tokenizer regressions, but they affect different models and different behavior (wrong token IDs vs encode/decode space stripping)." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper runtime failure is unrelated to the float8 storage lookup error." + "reason": "Different failures: Whisper model execution vs missing Float8 storage registration." }, { - "left": "issue:43632", - "right": "issue:44756", + "left": "issue:44291", + "right": "issue:44589", "accept": false, - "reason": "_is_hf_initialized handling and mmap OOM on Strix Halo are unrelated." + "reason": "Different bugs: init_empty_weights/_is_hf_initialized argument mismatch vs Float8 storage lookup failure." }, { - "left": "issue:44291", - "right": "issue:44589", + "left": "issue:43632", + "right": "issue:44756", "accept": false, - "reason": "Init-empty-weights constructor mismatch is unrelated to float8 storage resolution." + "reason": "Unrelated: _is_hf_initialized flag regression vs mmap/OOM on Strix Halo." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Both are Qwen VL training topics, but multi-GPU training and multi-image iterable dataset failures are different problems." + "reason": "Both Qwen2-VL/Qwen2.5-VL training-related, but multi-GPU training and multi-image prompt failure are separate problems." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Perplexity anomaly on GPT-OSS is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Different models and symptoms: high perplexity on gpt-oss vs Whisper batch_decode skipping special tokens." }, { "left": "issue:42796", "right": "issue:46710", "accept": false, - "reason": "Different regressions in different model families: Mistral tokenizer BPE markers vs DeepSeek accuracy after upgrade." + "reason": "Both are tokenizer/model quality regressions, but one is a tokenizer marker bug and the other is an accuracy regression after version upgrade." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "GGUF loading failure for Qwen2.5 and missing pad_token_id on Llama4Config are unrelated." + "reason": "Unrelated model-loading failures: GGUF load failure vs missing pad_token_id on Llama4Config." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to the iterable-dataset multi-image failure." + "reason": "Different generation mechanics: progressive generation with inputs_embeds/past_key_values vs finetuning with multiple images per prompt." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment bug is unrelated to Gemma4 FSDP2 shared_kv_states KeyError." + "reason": "Different bugs: ModelOutput key assignment vs Gemma4 FSDP2 shared_kv_states rebuild issue." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights _is_hf_initialized argument mismatch are distinct." + "reason": "Different code paths and errors: Qwen3-MoE sliding_window vs init_empty_weights/_is_hf_initialized mismatch." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Runtime special-token post-processor updates are related in theme, but these are different concrete failures and not clearly the same bug." + "reason": "Both involve special tokens, but one is runtime post-processor updates and the other is add_special_tokens behavior for a specific tokenizer." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "CUDA assert during auto device mapping is unrelated to the Mistral regex patch AttributeError." + "reason": "Different failures: Qwen3 device mapping cuda assert vs Mistral regex patch crashing on tokenizer object shape." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": true, + "reason": "Both describe the same packaging problem: installing the torch extra leaves the TensorFlow backend unavailable/half-installed." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Unrelated: GGUF loading regression vs CLIPTokenizer model_max_length behavior." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both are AutoTokenizer regressions, but one ignores repository tokenizer.json while the other returns wrong token IDs for Granite models." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "Different model families and failures: PaddleOCR-VL/vLLM loading vs TimesFM window_size AttributeError." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "Different regression classes: ModelOutput key handling vs non-persistent buffer initialization in v5." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Different domains: backend packaging for TensorFlow vs loading a specific multimodal model in vLLM." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The one accepted pair, 45216/45357, is a plausible Qwen3.5 save_pretrained duplicate family based on the titles, and the rest are reasonably treated as separate bugs or different failure modes." + "feedback": "Grounded and conservative overall. The cluster is indeed heterogeneous, and the two accepted pairs look like plausible same-bug duplicates: 45216/45357 are both Qwen3.5 save_pretrained regressions, and 42175/45542 are both TensorFlow-backend packaging issues. The canonical-issue choice is cautious and supported by the packet." }, "evaluator_used": true, "retried": false, @@ -4057,34 +4113,38 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-517", "nodes": [ + "issue:34634", + "issue:35707", "issue:40444", "issue:40990", - "issue:42175", + "issue:41553", + "issue:41720", "issue:42796", "issue:42915", - "issue:43296", + "issue:43262", "issue:43317", "issue:43377", "issue:43454", "issue:43482", "issue:43493", + "issue:43582", "issue:43606", "issue:43611", "issue:43644", "issue:43646", "issue:43701", + "issue:43746", "issue:43818", + "issue:43828", "issue:43856", "issue:43873", "issue:43994", "issue:44079", "issue:44265", "issue:44291", - "issue:44462", "issue:44466", "issue:44479", "issue:44509", - "issue:44534", "issue:44829", "issue:44877", "issue:45072", @@ -4093,11 +4153,8 @@ "issue:45381", "issue:45479", "issue:45491", - "issue:45538", - "issue:45542", "issue:45584", "issue:45701", - "issue:45812", "issue:45823", "issue:45834", "issue:46402", @@ -4106,248 +4163,238 @@ "issue:47217" ], "soft_pairs": [ - "issue:42175|issue:45542", - "issue:43482|issue:45538", - "issue:43296|issue:46821", - "issue:44462|issue:45812", - "issue:44079|issue:44534", - "issue:42175|issue:43296", - "issue:43317|issue:43856", "issue:44509|issue:46909", + "issue:43317|issue:43856", "issue:43611|issue:43646", "issue:43377|issue:45491", "issue:43644|issue:44079", "issue:43818|issue:46402", "issue:44265|issue:45072", - "issue:43818|issue:45357", "issue:40444|issue:45381", + "issue:43818|issue:45357", "issue:43493|issue:43994", "issue:45325|issue:45834", - "issue:44829|issue:45479", "issue:45325|issue:45381", + "issue:44829|issue:45479", "issue:43454|issue:44466", - "issue:44291|issue:44877", "issue:40444|issue:47217", + "issue:44291|issue:44877", "issue:42915|issue:43482", "issue:43701|issue:45325", "issue:43873|issue:45701", "issue:42796|issue:45381", "issue:43606|issue:44265", "issue:40990|issue:45584", - "issue:44479|issue:45823" + "issue:44479|issue:45823", + "issue:43828|issue:46821", + "issue:43262|issue:43746", + "issue:35707|issue:41720", + "issue:41553|issue:43582", + "issue:34634|issue:35707" ], - "prepared_review_unit_hash": "c31abd13925e8f5081f3a7b0948cdf39590ddc1d594ce2ef0d60200f2c5e77a5", + "prepared_review_unit_hash": "7559fcd6ca482a0ef65d9018297864b629d9a4540061571512b83d249853492a", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35928, - "estimated_input_tokens": 8982, - "estimated_eval_tokens": 18220 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36258, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated closed issues with a few likely duplicate pairs around tokenizer loading, SigLIP2 behavior, and Qwen2.5-VL position-id handling. No PRs are present.", - "confidence": 0.74, - "canonical_issue_reason": "issue:44462 is the broadest, most root-cause-oriented report in the set: it describes the underlying AutoTokenizer/tokenizer.json loading regression rather than a model-specific symptom.", + "summary": "This cluster is a loose mix of unrelated Transformers/model bugs, not a tight duplicate set. The only recurring pattern is broad v5 loading/init regressions, but the individual reports usually hit different code paths and models.", + "confidence": 0.37, + "canonical_issue_reason": "No true duplicate canonical exists; the closest broad anchor is issue 43611 because it represents the recurring v5 model-loading/init regression theme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44462 is the best triage anchor because it captures the general tokenizer-loading bug and is the clearest umbrella for the Granite-specific wrong-ID report.", + "best_issue_reason": "Issue 43611 is the broadest and most central representative of the set, so it is the best issue anchor even though it does not subsume the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42175", - "right": "issue:45542", - "accept": false, - "reason": "Related packaging/backend symptoms, but one is missing TensorFlow in the torch extra while the other is a TensorBoard-without-TensorFlow backend error; not the same bug." - }, - { - "left": "issue:43482", - "right": "issue:45538", - "accept": false, - "reason": "Different failures: Qwen2.5-GGUF loading under v5 versus CLIPTokenizer max-length handling." - }, - { - "left": "issue:43296", - "right": "issue:46821", - "accept": false, - "reason": "Unrelated model-specific load failure versus a TimesFM window_size AttributeError." - }, - { - "left": "issue:44462", - "right": "issue:45812", - "accept": true, - "reason": "Both describe the same AutoTokenizer/tokenizer.json regression, with 45812 appearing to be a model-specific manifestation on Granite models." - }, - { - "left": "issue:44079", - "right": "issue:44534", - "accept": false, - "reason": "Different subsystems: ModelOutput key assignment versus non-persistent buffer initialization." - }, - { - "left": "issue:42175", - "right": "issue:43296", + "left": "issue:44509", + "right": "issue:46909", "accept": false, - "reason": "No shared underlying bug; these are unrelated backend/install and vLLM model-loading issues." + "reason": "Both are documentation problems, but they concern different docs and different broken examples." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Both involve memory/offload themes, but one is a dequantized model load/offload failure and the other is Qwen3 MoE training memory usage." - }, - { - "left": "issue:44509", - "right": "issue:46909", - "accept": false, - "reason": "Both are documentation problems, but they affect different docs and different content." + "reason": "One is a loading/offload failure; the other is a training memory-usage report for a different model." }, { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both are v5 regressions, but one is about base_model_prefix model loading and the other about custom model initialization; too broad to merge." + "reason": "Both involve v5 model initialization/loading, but they are distinct failures with different triggers and fixes." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both concern batched/mixed-length behavior, but the root causes and models differ (padding mask support vs sliding-window all-padding NaNs)." + "reason": "Both are batching/padding-related, but they affect different models and different failure modes." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Different bugs: non-persistent buffer junk versus ModelOutput key handling." + "reason": "One is about non-persistent buffers in serialization; the other is ModelOutput key assignment logic." }, { "left": "issue:43818", "right": "issue:46402", "accept": false, - "reason": "Different model families and failures: Video-LLaVA tower wiring versus Sam3Model text-encoder loading." + "reason": "Different models and different bugs: missing video_tower temporal attention vs text encoder weights not loading." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "No shared code-path problem; torch.export/torch_compilable_check is unrelated to bfloat16 dtype mismatches." + "reason": "Export failure from torch_compilable_check is unrelated to the bf16 dtype-mismatch inference bug." }, { - "left": "issue:43818", - "right": "issue:45357", + "left": "issue:40444", + "right": "issue:45381", "accept": false, - "reason": "Both are multimodal model issues, but one is Video-LLaVA tower structure and the other is Qwen3.5 save_pretrained key serialization." + "reason": "Both are Qwen2.5-VL multimodal issues, but one is iterable-dataset multi-image finetuning and the other is video position-id computation." }, { - "left": "issue:40444", - "right": "issue:45381", + "left": "issue:43818", + "right": "issue:45357", "accept": false, - "reason": "Both touch Qwen2.5-VL multimodal inputs, but one is finetuning with multiple images per prompt and the other is video-position-id regression." + "reason": "Same broad family, but one is a Video-LLaVA architecture issue and the other is Qwen3.5 save_pretrained key handling." }, { "left": "issue:43493", "right": "issue:43994", - "accept": true, - "reason": "Same SigLIP2 problem: 43994 is the concrete bad-output symptom of the implementation discrepancy reported in 43493." + "accept": false, + "reason": "Both mention SigLIP2, but one is an implementation discrepancy and the other is a nonsensical-output report; not clearly the same bug." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Different model paths: Qwen2.5-VL rope-index position ids versus Kosmos2.5 OCR index errors." + "reason": "Both are index/position problems, but in different models and different input paths." }, { - "left": "issue:44829", - "right": "issue:45479", + "left": "issue:45325", + "right": "issue:45381", "accept": false, - "reason": "Different training pathologies with different triggers: flash_attention_3 regression versus single-label num_labels=1 zero-loss behavior." + "reason": "Both touch Qwen2.5-VL position handling, but one is still-image RoPE scaling and the other is video vision_position_ids." }, { - "left": "issue:45325", - "right": "issue:45381", - "accept": true, - "reason": "Both are Qwen2.5-VL v5.3.0 position-id regressions in the same multimodal code path; they plausibly reduce to one rope-index bug." + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both are training/loss regressions, but one is flash-attention-3-specific and the other is a single-label classification edge case." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Both mention lm_head weights, but one is a tie-weight initialization bug and the other is device-dependent serialization; not the same defect." + "reason": "Both involve lm_head/tied weights, but one is a model config/load bug and the other is a serialization inconsistency." }, { - "left": "issue:44291", - "right": "issue:44877", + "left": "issue:40444", + "right": "issue:47217", "accept": false, - "reason": "Different issues: init_empty_weights unexpectedly receiving _is_hf_initialized versus strict config rejecting granite_speech." + "reason": "Both are VL input handling issues, but one is dataset/multi-image finetuning and the other is image-placeholder-in-text parsing." }, { - "left": "issue:40444", - "right": "issue:47217", + "left": "issue:44291", + "right": "issue:44877", "accept": false, - "reason": "Different Qwen2.5-VL failure modes: finetuning with multiple images per prompt versus image-placeholder-in-user-text handling." + "reason": "Both are v5 loading regressions, but they stem from different strictness/argument-handling paths." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Unrelated Qwen3 MoE FP8 config failure and Qwen2.5-GGUF loading failure." + "reason": "Different quantization/loading failures: FineGrainedFP8Config vs GGUF loading under v5." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to Qwen2.5-VL position-id computation." + "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen2.5-VL RoPE/position-id handling." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Offloading/quantization behavior is unrelated to tokenizer-version changing tokenization." + "reason": "Quantization/offloading behavior is unrelated to a tokenization-version regression." }, { "left": "issue:42796", "right": "issue:45381", "accept": false, - "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video position ids are different bugs." + "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL video position IDs." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU-offload device mismatch for bark-small is unrelated to torch.export and torch_compilable_check." + "reason": "CPU-offload device mismatch and torch.export/compilability are different code paths." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Different model/task failures: GPT-OSS perplexity on WikiText-2 versus Whisper empty-transcription generation after align_special_tokens." + "reason": "Perplexity regression on gpt-oss-20b is unrelated to Whisper empty-transcription generation failure." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Both are v5 regressions, but they affect different models and different mechanisms: video-input handling versus Gemma4 PLE device mismatch." + "reason": "Both are inference regressions, but one is Qwen video-input handling and the other is Gemma4 device-map forward placement." + }, + { + "left": "issue:43828", + "right": "issue:46821", + "accept": false, + "reason": "An autocast dtype mismatch is unrelated to the TimesFM window_size AttributeError." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio processor chat-template sample-rate defaults and GraniteSpeech PEFT checkpoint loading are different issues." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to the Qwen3 auto-device-map CUDA assert." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Bad AutoTokenizer error messaging and an AppleSilicon allocator TypeError are unrelated." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are different bugs." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately conservative: it notes only a few likely duplicate issue pairs, and the accepted pairs are reasonably supported by the issue titles. No PRs are present, so there is no mergeability concern." - }, - "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 }, @@ -4364,35 +4411,33 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-517", "nodes": [ - "issue:34634", - "issue:35707", + "issue:26413", + "issue:28530", "issue:36331", "issue:40990", - "issue:41553", "issue:41720", "issue:42617", "issue:42796", "issue:42915", "issue:42947", + "issue:43065", "issue:43257", - "issue:43262", "issue:43317", "issue:43322", "issue:43377", + "issue:43450", "issue:43454", "issue:43550", "issue:43572", - "issue:43582", "issue:43606", "issue:43720", - "issue:43746", "issue:43749", "issue:43819", - "issue:43828", "issue:43873", "issue:44186", "issue:44265", "issue:44387", + "issue:44466", "issue:44534", "issue:44568", "issue:44610", @@ -4401,6 +4446,7 @@ "issue:44977", "issue:44991", "issue:45072", + "issue:45357", "issue:45701", "issue:45704", "issue:45759", @@ -4411,31 +4457,27 @@ "issue:46209", "issue:46305", "issue:46315", + "issue:46402", "issue:46821" ], "soft_pairs": [ - "issue:41553|issue:43582", - "issue:43828|issue:46821", - "issue:43262|issue:43746", - "issue:35707|issue:41720", - "issue:34634|issue:35707", - "issue:41720|issue:45823", "issue:44186|issue:44265", + "issue:41720|issue:45823", "issue:43606|issue:43819", "issue:43454|issue:43550", "issue:43257|issue:45923", "issue:45823|issue:46821", - "issue:44568|issue:45920", "issue:42796|issue:46209", "issue:43377|issue:44610", + "issue:44568|issue:45920", "issue:43454|issue:45072", "issue:42617|issue:42915", "issue:40990|issue:42947", "issue:46305|issue:46315", - "issue:43572|issue:45759", "issue:43377|issue:44898", "issue:43454|issue:43720", "issue:43749|issue:44534", + "issue:43572|issue:45759", "issue:43317|issue:43873", "issue:36331|issue:42617", "issue:41720|issue:43322", @@ -4445,228 +4487,226 @@ "issue:41720|issue:44977", "issue:44387|issue:45704", "issue:40990|issue:45701", - "issue:41720|issue:42617" + "issue:41720|issue:42617", + "issue:43450|issue:45701", + "issue:43065|issue:46402", + "issue:44466|issue:45357", + "issue:26413|issue:28530" ], - "prepared_review_unit_hash": "4d83363fb6d1b1d0eaff776649f8a6a0407c7edab37a3a48364165855e63038e", + "prepared_review_unit_hash": "3c0672571603c07df8342419f94e6fa0b05ebe1f6a410e7a45b250390454f6fc", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37314, - "estimated_input_tokens": 9329, - "estimated_eval_tokens": 18914 + "soft_pair_count": 31, + "serialized_chars": 37063, + "estimated_input_tokens": 9266, + "estimated_eval_tokens": 18788 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most pairs are only loosely similar by subsystem wording (loading, tokenization, device_map, offloading, dtype), but they describe different concrete failures, so none should be merged as duplicates.", - "confidence": 0.34, - "canonical_issue_reason": "issue:43873 is the broadest and most central open issue in the cluster, with the most discussion/inbound references and the closest umbrella theme around offloading/quantization.", + "summary": "This cluster is mostly a set of unrelated closed issues that only share broad subsystem similarity (loading, tokenization, offloading, quantization, training-loop regressions). I found no true duplicate pairs among the soft candidates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 43317 is the clearest representative of the shared theme around model loading/offload regressions, with a concrete and actionable failure mode; however, the cluster as a whole is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the best cluster representative for triage because it captures the dominant offloading/device-loading theme and is still open for follow-up.", + "best_issue_reason": "Issue 43317 is the strongest standalone issue here because it describes a specific, user-relevant regression in a common code path (`device_map=auto` + offload) without being overly model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:41553", - "right": "issue:43582", - "accept": false, - "reason": "Different problems: bad tokenizer error message for Voxtral vs AppleSilicon TypeError in caching_allocator_warmup." - }, - { - "left": "issue:43828", - "right": "issue:46821", - "accept": false, - "reason": "Unrelated: autocast dtype mismatch in Phi-tiny-MoE vs TimesFM window_size AttributeError." - }, - { - "left": "issue:43262", - "right": "issue:43746", - "accept": false, - "reason": "Different areas and failures: audio chat template sampling-rate default vs GraniteSpeech PEFT checkpoint loading." - }, - { - "left": "issue:35707", - "right": "issue:41720", - "accept": false, - "reason": "Different bugs: progressive generation with inputs_embeds/past_key_values vs Qwen3 auto device mapping cuda assert." - }, - { - "left": "issue:34634", - "right": "issue:35707", + "left": "issue:44186", + "right": "issue:44265", "accept": false, - "reason": "Unrelated: BarkProcessor voice_preset issue vs generation with inputs_embeds/past_key_values." + "reason": "Both are export/processing failures, but one is a tokenizer crash on NER batching while the other is `torch.export` failing on `torch_compilable_check`; different code paths and fixes." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Both involve device_map auto, but the concrete failures differ (cudaErrorAssert vs device mismatch in Gemma4 PLE)." - }, - { - "left": "issue:44186", - "right": "issue:44265", - "accept": false, - "reason": "Tokenizer NER/padding crash is not the same as torch.export export failure with torch_compilable_check." + "reason": "Both mention auto device mapping, but the concrete bugs differ: CUDA assert on Qwen3 vs a Gemma4 PLE device mismatch during forward." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "Different model/code-path issues: bark-small CPU offload device mismatch vs DAC.from_latents/STE mismatch." + "reason": "One is a CPU offload device-mismatch bug for `suno/bark-small`; the other is a DAC latent decoding mismatch caused by missing STE. Same broad area, different defect." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "Different models and failure modes: AyaVision lm_head tying vs Bamba torch.compile with SDPA." + "reason": "AyaVision lm_head tying failure and Bamba torch.compile+SDPA failure are unrelated model/runtime bugs." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Different problems: Qwen3 MoE weight conversion under accelerate+deepspeed vs missing supports_gradient_checkpointing flag." + "reason": "Qwen3 MoE loading with accelerate+deepspeed and a missing `supports_gradient_checkpointing` flag in Nemotron are separate issues." }, { "left": "issue:45823", "right": "issue:46821", "accept": false, - "reason": "Different bugs: Gemma4 PLE device mismatch vs TimesFM window_size attribute error." - }, - { - "left": "issue:44568", - "right": "issue:45920", - "accept": false, - "reason": "Both are tokenizer regressions, but one is missing BOS/EOS on a specific tokenizer while the other is wrong token IDs across several models." + "reason": "Gemma4 device mismatch under `device_map=\"auto\"` and TimesFM `window_size` AttributeError are different models and different failure modes." }, { "left": "issue:42796", "right": "issue:46209", "accept": false, - "reason": "Tokenizer BPE-marker handling is unrelated to FP8 inference failure on B200." + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are unrelated." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI batched-vs-single padding-mask bug is unrelated to OmDet-Turbo processor image-size mismatch." + "reason": "MIMI batch-vs-single padding-mask inconsistency and OmDet-Turbo processor output-size mismatch affect different components and bugs." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "One is a single tokenizer missing BOS/EOS insertion; the other is a broader v5 regression producing wrong token IDs across many tokenizers." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "Different failures: lm_head weight tying/generation garbage vs dtype mismatch in SwitchTransformers/TimmWrapperModel." + "reason": "AyaVision lm_head tying/garbage text and dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "3d_parallel.py execution failure is not the same as Qwen3Moe FineGrainedFP8Config loading." + "reason": "A `3d_parallel.py` runtime failure is not the same as Qwen3Moe failing with `FineGrainedFP8Config`." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Perplexity regression on gpt-oss-20b is unrelated to LoRA gradient checkpointing ineffectiveness." + "reason": "High perplexity on gpt-oss and ineffective gradient checkpointing with PEFT LoRA are different symptoms and code paths." }, { "left": "issue:46305", "right": "issue:46315", "accept": false, - "reason": "Different tokenizer/windowing regressions: streaming mask off-by-one vs convert_tokens_to_ids performance." - }, - { - "left": "issue:43572", - "right": "issue:45759", - "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is not the same as composite Qwen multimodal text_config unwrapping." + "reason": "MoonshineStreaming mask off-by-one and tokenizer conversion slowdown are unrelated regressions." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Both involve model processing, but one is padding-mask output divergence and the other is Perceiver resolution handling." + "reason": "MIMI batching/padding-mask behavior and Perceiver interpolation on non-default resolution are distinct model bugs." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "AyaVision lm_head tying and BitNet packed-weight unpacking are different loading/generation issues." + "reason": "lm_head tying issues are unrelated to packed-weight unpacking during accelerate loading." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "FSDP CPU RAM efficient loading is not the same underlying bug as non-persistent buffers being junk in v5." + "reason": "FSDP CPU RAM efficient loading and non-persistent buffer junk are both v5-loading related, but they are not the same concrete bug." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing `pad_token_idx` in a config and failure to unwrap `text_config` in composite Qwen configs are different config-loading problems." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Closely related theme, but different concrete issues: dequantized model load failure with gpu+cpu offload vs generic quantization offloading behavior." + "reason": "Both involve offloading/loading, but 43317 is a specific `device_map=auto` dequantized-model failure while 43873 is a broader quantization/offloading behavior report." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Unexpected num_items_in_batch argument in CustomTrainer.compute_loss is unrelated to 3d_parallel.py failure." + "reason": "`compute_loss()` kwarg incompatibility and a `3d_parallel.py` failure are unrelated training/script issues." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Qwen3 device_map auto cuda assert and Llava Next segmentation fault are different code paths and failures." + "reason": "Qwen3 auto device map assert and Llava Next segmentation fault are different models and failure mechanisms." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "Segfault loading Llava Next models is unrelated to sinq quantized save/load breakage." + "reason": "A load-time segfault for Llava Next and broken quantized save/load after `sinq` are different bugs." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict granite_speech config loading and EMBEDDIA/est-roberta tokenizer loading are different regressions." + "reason": "Strict config blocking `granite_speech` and tokenizer loading for `est-roberta` are separate config/tokenizer issues." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Trainer API signature breakage is unrelated to Qwen3Moe FP8 config failure." + "reason": "The `compute_loss` signature mismatch is unrelated to Qwen3Moe `FineGrainedFP8Config` failures." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Different bugs: auto device mapping assert vs Qwen3.5 flash-attention generation issue." + "reason": "Auto device-map CUDA assert in Qwen3 and flash-attention generation issues in Qwen3.5 are distinct runtime bugs." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "CUDA reserved memory OOM under int4 quantization is not the same as T5 using apex.FusedRMSNorm memory leak." + "reason": "CUDA reserved-memory OOM under int4 quantization and an apex RMSNorm memory leak are not the same problem." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "High perplexity on gpt-oss-20b is unrelated to version-dependent tokenization changes." + "reason": "High perplexity on a specific model and version-dependent tokenization changes are different regressions." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Different failures: Qwen3 auto device mapping assert vs inability to run 3d_parallel.py." + "reason": "A model-loading CUDA assert is not the same as a `3d_parallel.py` execution failure." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape errors and tokenizer-version behavior changes are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and missing Sam3 text-encoder weights are different issues in different parts of the model." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "lm_head serialization inconsistency and incorrect visual-encoder keys on save_pretrained are distinct save/load regressions." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Checkpoint-resume epoch iterator failure and early-stopping metric detection are different trainer bugs." } ] }, @@ -4690,18 +4730,16 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-517", "nodes": [ - "issue:26413", "issue:28530", "issue:33666", "issue:36010", + "issue:36331", "issue:41720", "issue:42617", "issue:42915", - "issue:43065", "issue:43295", "issue:43317", "issue:43404", - "issue:43450", "issue:43531", "issue:43550", "issue:43572", @@ -4716,10 +4754,12 @@ "issue:44186", "issue:44423", "issue:44451", - "issue:44466", + "issue:44464", "issue:44488", "issue:44610", + "issue:44671", "issue:44734", + "issue:44743", "issue:44756", "issue:44898", "issue:44991", @@ -4731,7 +4771,6 @@ "issue:45701", "issue:45823", "issue:46291", - "issue:46402", "issue:46612", "issue:46665", "issue:46821", @@ -4740,244 +4779,237 @@ "issue:47020" ], "soft_pairs": [ - "issue:43450|issue:45701", - "issue:43065|issue:46402", - "issue:44466|issue:45357", "issue:43881|issue:44488", - "issue:26413|issue:28530", "issue:43873|issue:44756", "issue:44186|issue:45072", - "issue:45685|issue:46821", "issue:46612|issue:46858", + "issue:45685|issue:46821", "issue:43295|issue:45701", - "issue:43577|issue:43708", "issue:43550|issue:43720", - "issue:43646|issue:46665", + "issue:43577|issue:43708", "issue:36010|issue:42617", + "issue:43646|issue:46665", "issue:44451|issue:47020", "issue:43550|issue:43606", "issue:44186|issue:44610", "issue:42915|issue:43317", - "issue:46821|issue:46869", "issue:43404|issue:45357", + "issue:46821|issue:46869", "issue:44610|issue:44898", "issue:28530|issue:33666", "issue:41720|issue:45161", "issue:44734|issue:46858", "issue:43531|issue:43572", "issue:44008|issue:45823", - "issue:42915|issue:46291", "issue:44423|issue:45406", - "issue:43881|issue:44991" + "issue:42915|issue:46291", + "issue:36331|issue:41720", + "issue:44464|issue:46858", + "issue:43881|issue:44991", + "issue:44671|issue:44743" ], - "prepared_review_unit_hash": "a2091eb59484f948f0b4bd0afdf9502a135a356eae2b6454a546798c26267ac0", + "prepared_review_unit_hash": "bc1e1547e04ae46f7fb9f6491d5cb078f3772ea06c0e1d2ae32575bd9bc9eda7", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36452, - "estimated_input_tokens": 9113, - "estimated_eval_tokens": 18482 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35575, + "estimated_input_tokens": 8894, + "estimated_eval_tokens": 18044 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a noisy semantic mix of unrelated Transformers issues. The strongest common thread is loading/offload/quantization/device-map regressions, but the soft pairs are too broad or model-specific to merge. No PRs are present; the most representative issue is the open offloading/quantization report.", - "confidence": 0.83, - "canonical_issue_reason": "issue:43873 is the clearest hub for the dominant theme in this set: offloading/quantization/device placement failures. It is open, has the most discussion/inbound refs, and best represents the recurring load/offload class of bugs.", + "summary": "The set is mostly a loose similarity cluster around Transformers regressions in loading, generation, cache, and quantization, but the soft pairs do not look like true duplicates of the same underlying bug. No PRs are present.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43646 is the broadest representative of the cluster: a Transformers 5.0 regression breaking custom model initialization and related downstream loading/custom-code failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is also the best overall issue for triage because it has the strongest signal/engagement and captures the broadest recurring failure mode in the cluster.", + "best_issue_reason": "issue:43646 is the best standalone exemplar because it captures the widest recurring theme in the cluster (breaking API/init behavior after version changes) without being overly model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43450", - "right": "issue:45701", - "accept": false, - "reason": "Different bugs: video batch shape handling vs tokenizer behavior changes. Only broad preprocessing similarity." - }, - { - "left": "issue:43065", - "right": "issue:46402", - "accept": false, - "reason": "Different models and failures: dummy Conv2d placeholder vs missing text encoder weights." - }, - { - "left": "issue:44466", - "right": "issue:45357", - "accept": false, - "reason": "Both involve serialization, but one is tied lm_head weights and the other incorrect visual encoder keys; not the same concrete bug." - }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "Both are loading failures, but for different models and likely different root causes." - }, - { - "left": "issue:26413", - "right": "issue:28530", - "accept": false, - "reason": "Trainer loop bugs, but one is checkpoint resume/epoch iterator and the other is early stopping metric detection." + "reason": "Both are model loading failures, but they involve different models and likely different root causes." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Both touch memory/offload behavior, but one is quantization offloading and the other is mmap/OOM on Strix Halo." + "reason": "Both touch offload/memory behavior, but one is quantization offloading and the other is mmap/OOM on a specific platform; not the same bug." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Unrelated failures: tokenizer crashes vs dtype mismatch in inference." - }, - { - "left": "issue:45685", - "right": "issue:46821", - "accept": false, - "reason": "Different subsystems and symptoms: MPS histogram/int error vs TimesFM window_size attribute error." + "reason": "Tokenizer/NER padding crash vs bfloat16 dtype mismatch are different failure modes." }, { "left": "issue:46612", "right": "issue:46858", "accept": false, - "reason": "Both concern cache code, but one is beam-search reorder skipping and the other is static-cache slice arithmetic on non-CUDA backends." + "reason": "Both involve cache internals, but one is beam-search reorder and the other is seq-length tensor arithmetic in static cache; different code paths." }, { - "left": "issue:43295", - "right": "issue:45701", + "left": "issue:45685", + "right": "issue:46821", "accept": false, - "reason": "Both mention tokenizer/processor behavior, but one is a regression in processor.tokenizer access and the other is version-dependent tokenization differences." + "reason": "MPS histogram/int error and TimesFM window_size AttributeError are unrelated bugs." }, { - "left": "issue:43577", - "right": "issue:43708", + "left": "issue:43295", + "right": "issue:45701", "accept": false, - "reason": "Different areas: BLIP2 dtype propagation vs Trainer resume/max_steps calculation." + "reason": "Both are version-related regressions, but they concern different API breakages and behaviors." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "Different code paths: torch.compile SDPA failure vs accelerate loading/unpacking packed BitNet weights." + "reason": "torch.compile/SDPA failure and packed-weight loading failure are unrelated." }, { - "left": "issue:43646", - "right": "issue:46665", + "left": "issue:43577", + "right": "issue:43708", "accept": false, - "reason": "Both relate to model construction/loading, but not the same bug or code path." + "reason": "Model dtype loading bug vs resume_from_checkpoint max_steps miscalculation are different issues." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "Unrelated issues: GenerationMixin import breakage vs inability to run a parallel script." + "reason": "ImportError for GenerationMixin and inability to run a script are not the same underlying bug." + }, + { + "left": "issue:43646", + "right": "issue:46665", + "accept": false, + "reason": "Both mention custom model/from_pretrained behavior, but the reported failures and code paths are too different to merge." }, { "left": "issue:44451", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer-related, but one is loading a specific model and the other is a Unigram constructor type mismatch." + "reason": "Different tokenizer-loading failures on different models; no concrete shared root cause is established." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Different problems: SDPA/torch.compile failure vs CPU offload device mismatch." + "reason": "torch.compile with SDPA and CPU offload device mismatch are separate runtime bugs." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Different image/preprocessing bugs with different failure modes; not the same underlying issue." + "reason": "Tokenizer crash and image-size mismatch are both preprocessing-related, but not the same defect." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Both involve quantization/offload/loading, but one is Qwen3Moe FP8 config and the other is dequantized model loading with device_map=auto." + "reason": "FP8 fine-grained config failure and dequantized model offload failure are distinct quantization/loading problems." }, { - "left": "issue:46821", - "right": "issue:46869", + "left": "issue:43404", + "right": "issue:45357", "accept": false, - "reason": "Both are device-map-ish model failures, but on different models and unrelated error paths." + "reason": "lm_head tying bug and save_pretrained visual-key regression are different state/serialization issues." }, { - "left": "issue:43404", - "right": "issue:45357", + "left": "issue:46821", + "right": "issue:46869", "accept": false, - "reason": "Both are weight serialization/tie issues in multimodal models, but they affect different tensors and different model families." + "reason": "TimesFM argument handling and YOLO multi-GPU device-map failure are unrelated." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both involve image resolution expectations, but one is processor output size and the other is Perceiver behavior with interpolate_pos_encoding." + "reason": "Both concern image resolution handling, but they affect different models and different preprocessing/forward logic." }, { "left": "issue:28530", "right": "issue:33666", "accept": false, - "reason": "Completely different problems: early stopping config vs Qwen2-VL multi-GPU training." + "reason": "Early stopping metric configuration and Qwen2-VL multi-GPU training are unrelated." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Both are multi-device/model-parallel topics, but the failures and models are unrelated." + "reason": "Both are parallelism/device-map problems, but one is a cuda assert and the other is tensor parallelism support; not the same bug." }, { "left": "issue:44734", "right": "issue:46858", "accept": false, - "reason": "Both mention cache arithmetic, but one is serve/KV continuation indexing and the other is static cache length on non-CUDA backends." + "reason": "Both involve cache indexing, but one is serve KV continuation and the other is static-cache prefill arithmetic on non-CUDA backends." }, { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Different regressions in different models/configs; only superficial similarity around configuration problems." + "reason": "Qwen3-MoE sliding_window behavior and StableLmConfig pad_token_idx regression are unrelated." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Same broad Gemma family, but the bugs differ: variable name collision vs device-mapping mismatch." + "reason": "Gemma3n forward variable collision and Gemma4 PLE device mismatch are different forward-pass bugs." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both hit transformers serve with multimodal models, but the crash causes and failing paths are different." }, { "left": "issue:42915", "right": "issue:46291", "accept": false, - "reason": "Unrelated: FP8 quantized model failure vs pipeline import dependency/version issue." + "reason": "FP8 model failure and kernels package import/version requirement are unrelated." }, { - "left": "issue:44423", - "right": "issue:45406", + "left": "issue:46612", + "right": "issue:46858", "accept": false, - "reason": "Different service crashes with different attributes and model/processor classes." + "reason": "Same general cache area, but distinct defects and different observable failures." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Trainer compute_loss signature change and multi-GPU cuda assert are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:46858", + "accept": false, + "reason": "Both mention generation/cache and compiled forward, but the reported bugs are not the same concrete issue." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "Both are load failures, but one is a model-specific load error and the other is tokenizer compatibility for a different checkpoint." + "reason": "Different tokenizer/model loading failures with no shared root cause." }, { - "left": "issue:467?", - "right": "issue:46858", + "left": "issue:44671", + "right": "issue:44743", "accept": false, - "reason": "Not applicable." + "reason": "Masked LM prediction regression and recurrent-state reset in cache are different model-specific bugs." } ] }, @@ -5009,13 +5041,13 @@ "issue:34634", "issue:35707", "issue:36032", - "issue:36331", "issue:41720", "issue:42490", "issue:42915", "issue:43316", "issue:43322", "issue:43329", + "issue:43388", "issue:43425", "issue:43452", "issue:43526", @@ -5031,30 +5063,27 @@ "issue:43883", "issue:44186", "issue:44361", - "issue:44464", "issue:44492", "issue:44509", "issue:44534", "issue:44589", - "issue:44671", - "issue:44743", "issue:44913", + "issue:45059", + "issue:45406", + "issue:45464", "issue:45632", "issue:45684", "issue:45701", "issue:45704", "issue:45759", + "issue:45859", "issue:46665", "issue:46821", - "issue:46858", "issue:47020", "issue:47145" ], "soft_pairs": [ - "issue:36331|issue:41720", - "issue:44464|issue:46858", "issue:43526|issue:43883", - "issue:44671|issue:44743", "issue:45632|issue:45684", "issue:44186|issue:44361", "issue:43425|issue:46821", @@ -5069,32 +5098,35 @@ "issue:43540|issue:43782", "issue:43531|issue:45759", "issue:44492|issue:44509", - "issue:43452|issue:47020", "issue:43550|issue:43653", + "issue:43452|issue:47020", "issue:44534|issue:46665", "issue:41720|issue:43746", "issue:43606|issue:43720", "issue:42490|issue:43329", - "issue:43531|issue:46821", "issue:34567|issue:35707", + "issue:43531|issue:46821", "issue:34634|issue:41720", + "issue:45406|issue:45464", + "issue:43606|issue:43653", + "issue:43388|issue:45059", "issue:43531|issue:44589", - "issue:43606|issue:43653" + "issue:45759|issue:45859" ], - "prepared_review_unit_hash": "be0d7b6b85c74be4ccaa3a8820aeca1eba8b6dbeac73feabc521c56ed98f67ea", + "prepared_review_unit_hash": "331842db69c58689eba96bc0fa74a6e9f949ecf4b4b73a6d5ef81b3c3071d0fb", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36410, + "serialized_chars": 36411, "estimated_input_tokens": 9103, "estimated_eval_tokens": 18462 }, @@ -5103,186 +5135,186 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are a heterogeneous mix of unrelated bug reports across tokenizers, multimodal models, training, caching, and quantization. None of the soft pairs describe the same concrete underlying bug/change closely enough to merge, so all candidate duplicate edges should be rejected.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "This cluster is a grab bag of unrelated Transformers issues across tokenizers, multimodal processors, generation, loading/offloading, configs, and docs. None of the soft pairs look like the same underlying bug/change, so there is no credible duplicate consolidation here.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out: the items span many distinct subsystems and failure modes, so any representative would be misleading.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No strong global best issue: even the broadest reports are specific to one model/path and do not represent the cluster as a whole.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:36331", - "right": "issue:41720", - "accept": false, - "reason": "Different failures and code paths: trainer custom loss signature vs Qwen3 auto device-mapping CUDA assert." - }, - { - "left": "issue:44464", - "right": "issue:46858", - "accept": false, - "reason": "Both involve generation/cache, but one is chunked generation with compiled forward while the other is a static-cache seq-length tensor arithmetic bug." - }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "Unrelated subsystems: BeitImageProcessorFast label reduction vs Molmo loading tied-weights attribute error." - }, - { - "left": "issue:44671", - "right": "issue:44743", - "accept": false, - "reason": "CamemBERT masked-LM regression and Qwen modular recurrent-state reset are different model-specific bugs." + "reason": "Different bugs: BeitImageProcessorFast label reduction vs MolmoForCausalLM missing all_tied_weights_keys." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "One is a cache-path collision for trust_remote_code; the other is a permissions propagation problem in save_pretrained." + "reason": "Different loading problems: local cache-path collision vs save_pretrained permission propagation." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Both are tokenizer/task errors, but they affect different tokenizers and different failure mechanisms." + "reason": "Different tokenizer failures in different models and code paths (LayoutLMv2 NER padding vs MLuke task AttributeError)." }, { "left": "issue:43425", "right": "issue:46821", "accept": false, - "reason": "Torch version incompatibility is not the same as a TimesFM window_size AttributeError." + "reason": "Torch 2.10 compatibility issue is unrelated to TimesFM window_size AttributeError." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Early stopping metric lookup and void segmentation-map processing are unrelated." + "reason": "Training early-stopping metric selection vs image processor void-segmentation handling are unrelated." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "API discrepancy in Gemma3TextConfig is unrelated to GPTNeoX rotary_pct reloading behavior." + "reason": "Different config/API bugs: Gemma3TextConfig discrepancy vs GPTNeoX rotary_pct reload regression." }, { "left": "issue:45701", "right": "issue:47145", "accept": false, - "reason": "Tokenization-version behavior changes are unrelated to FP8 checkpoint weight_scale loading corruption." + "reason": "Broad tokenization-version behavior vs FP8 compressed-tensors checkpoint loading corruption; not the same defect." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Llava Next segmentation fault are different model-loading bugs." + "reason": "Different models and symptoms: Qwen3Moe FineGrainedFP8Config vs Llava Next segfault." }, { "left": "issue:33357", "right": "issue:35707", "accept": false, - "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are distinct problems." + "reason": "MacOS CLIP bus error vs generation with inputs_embeds/past_key_values are unrelated." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Tokenizer inconsistency between slow/fast and a T5 add_special_tokens load conflict are not the same bug." + "reason": "Tokenizer inconsistency across slow/fast implementations is not the same as T5 add_special_tokens name conflict." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Image segmentation-map processing and BarkProcessor voice_preset handling are unrelated." + "reason": "Image segmentation-map processing vs BarkProcessor voice_preset are unrelated processor bugs." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantization offloading misbehavior and T5 apex.FusedRMSNorm memory leak do not share the same code path." + "reason": "Quantized offloading behavior vs T5 apex RMSNorm memory leak are different code paths and root causes." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Video-input ValueError in Qwen3OmniMoe is not the same as Qwen3VL weight_only from_pretrained failure." + "reason": "Different multimodal/Qwen bugs: video input ValueError vs weight_only loading error." }, { "left": "issue:43531", "right": "issue:45759", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE and config unwrapping for composite multimodal models are different issues." + "reason": "Both Qwen-related, but one is sliding_window behavior and the other is from_config text_config unwrapping; different bugs." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A cache-strategy typo is unrelated to outdated docs mentioning removed pipeline tasks." + "reason": "A typo in cache strategies is not the same as stale docs for removed pipeline tasks." }, { - "left": "issue:43452", - "right": "issue:47020", + "left": "issue:43550", + "right": "issue:43653", "accept": false, - "reason": "gguf_file from_pretrained breakage is not the same as XLMRobertaTokenizer passing a dict to Unigram." + "reason": "torch.compile+SDPA failure in Bamba is unrelated to BigBirdTokenizer special-token registration." }, { - "left": "issue:43550", - "right": "issue:43653", + "left": "issue:43452", + "right": "issue:47020", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba is unrelated to BigBirdTokenizer special-token registration." + "reason": "Different tokenizer/model loading issues: gguf_file handling vs XLMRobertaTokenizer Unigram input type." }, { "left": "issue:44534", "right": "issue:46665", "accept": false, - "reason": "Non-persistent buffer corruption in v5 is distinct from from_pretrained behavior inside Python classes." + "reason": "Non-persistent buffer junk in v5 vs from_pretrained behavior inside Python classes are different bugs." }, { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Qwen3 CUDA assert during device mapping and GraniteSpeech PEFT local checkpoint loading are separate failures." + "reason": "Qwen3 auto device-mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading are unrelated." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "CPU offload device mismatch and packed-weight unpacking during accelerate loading are different bugs." + "reason": "CPU offload device mismatch for Bark-small vs BitNet packed-weight unpacking during accelerate loading are different issues." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "Trailing-slash local model import warning is unrelated to the undefined-variable multimodal video branch bug." + "reason": "Local-path trailing slash import warning is unrelated to undefined video-branch variables in multimodal token counting." }, { - "left": "issue:43531", - "right": "issue:46821", + "left": "issue:34567", + "right": "issue:35707", "accept": false, - "reason": "Both mention model-specific runtime behavior, but the failures and affected paths are different." + "reason": "TrainerState token counter not updating vs progressive generation with inputs_embeds/past_key_values are different problems." }, { - "left": "issue:34567", - "right": "issue:35707", + "left": "issue:43531", + "right": "issue:46821", "accept": false, - "reason": "TrainerState token-count tracking and progressive generation with cache are unrelated." + "reason": "Both mention windowing, but Qwen3-MoE sliding_window and TimesFM window_size are different model-specific bugs." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "Bark voice_preset loading and Qwen3 device mapping CUDA assert are different model bugs." + "reason": "BarkProcessor voice_preset failure and Qwen3 device-mapping CUDA assert are unrelated." }, { - "left": "issue:43531", - "right": "issue:44589", + "left": "issue:45406", + "right": "issue:45464", "accept": false, - "reason": "Qwen3 sliding-window behavior and missing Float8 storage handling are unrelated." + "reason": "Gemma4Processor missing _tokenizer in serve is not the same as Qwen3.5 streaming chat/completions failure." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Bark CPU offload mismatch and BigBirdTokenizer mask-token registration are not the same underlying issue." + "reason": "CPU offload device mismatch and BigBirdTokenizer decode/special-token issues are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "gather_for_metrics label truncation and SAM3 text/bbox odd behavior are different bugs." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug is unrelated to missing Float8 storage support." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "from_config text_config unwrapping and CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding are different issues." } ] }, @@ -5312,25 +5344,28 @@ "issue:41720", "issue:42886", "issue:43381", - "issue:43388", "issue:43421", "issue:43493", "issue:43526", "issue:43540", "issue:43644", + "issue:43653", "issue:43746", "issue:43844", "issue:43873", "issue:43881", "issue:43927", + "issue:43994", "issue:44062", "issue:44077", "issue:44368", "issue:44451", + "issue:44464", "issue:44514", "issue:44545", "issue:44610", "issue:44625", + "issue:44671", "issue:44843", "issue:44898", "issue:44918", @@ -5338,40 +5373,34 @@ "issue:44987", "issue:44991", "issue:45042", - "issue:45059", "issue:45072", "issue:45357", - "issue:45406", - "issue:45464", "issue:45479", "issue:45704", "issue:45759", - "issue:45859", "issue:45923", "issue:46050", "issue:46402", "issue:46655", "issue:46665", "issue:46693", - "issue:46858" - ], + "issue:46858", + "issue:47050" + ], "soft_pairs": [ - "issue:43388|issue:45059", - "issue:45406|issue:45464", - "issue:45759|issue:45859", "issue:44451|issue:44991", "issue:43540|issue:46858", - "issue:44077|issue:44625", "issue:45357|issue:46402", + "issue:44077|issue:44625", "issue:43644|issue:46665", "issue:44987|issue:45042", "issue:44062|issue:44451", "issue:44898|issue:45072", "issue:43381|issue:43844", "issue:44625|issue:45479", - "issue:43526|issue:43881", "issue:39692|issue:44610", "issue:41720|issue:45704", + "issue:43526|issue:43881", "issue:44368|issue:45357", "issue:43381|issue:45704", "issue:44918|issue:44936", @@ -5384,216 +5413,219 @@ "issue:42886|issue:44843", "issue:30064|issue:36010", "issue:44514|issue:44545", - "issue:39692|issue:43493" + "issue:39692|issue:43493", + "issue:43421|issue:43653", + "issue:43994|issue:44671", + "issue:44464|issue:47050" ], - "prepared_review_unit_hash": "ca4efeb0e3b73c3e13ac8095cb33c7c81b2c8a324e910fbf8adbb1c0e7e4c394", + "prepared_review_unit_hash": "abdbc1d57cb7011500813d29b33202fcacf519244a878b0461dcddad19130f38", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36709, - "estimated_input_tokens": 9178, - "estimated_eval_tokens": 18612 + "serialized_chars": 36678, + "estimated_input_tokens": 9170, + "estimated_eval_tokens": 18596 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are unrelated regressions or feature bugs across different subsystems. I only see two clear duplicate pairs: the Qwen2_5_VLProcessor batched-padding crash and the HF_HUB_OFFLINE tokenizer cache-load failure.", - "confidence": 0.92, - "canonical_issue_reason": "No single canonical issue covers the full set: these are mostly unrelated bugs spanning image processors, tokenizers, training, serving, and model loading. The only clear duplicate groups are 42886\u219444843 and 44514\u219444545.", + "summary": "This cluster is mostly heterogeneous issue reports, but two soft-duplicate pairs stand out: the Qwen2_5_VLProcessor batched apply_chat_template crash, and the HF_HUB_OFFLINE tokenizer/model_info offline-mode failure. The rest look like related-but-distinct bugs rather than the same underlying change.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 46693 is the strongest single anchor by activity and impact: it is open, has the highest discussion volume, and describes a broad performance regression affecting a major generation path.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue is suitable; the items do not form one underlying bug cluster, so any single issue would misrepresent the rest.", + "best_issue_reason": "Issue 46693 is the best overall representative because it is actively discussed, broadly scoped, and most likely to be the cluster\u2019s main user-facing problem rather than a narrow edge case.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43388", - "right": "issue:45059", - "accept": false, - "reason": "Different bugs: one is gather_for_metrics dropping tuple labels in the last batch, the other is a SAM3 PCS text/box behavior issue." - }, - { - "left": "issue:45406", - "right": "issue:45464", - "accept": false, - "reason": "Different failure paths: Gemma4Processor missing _tokenizer in transformers serve vs Qwen3.5 streaming inference crash." - }, - { - "left": "issue:45759", - "right": "issue:45859", - "accept": false, - "reason": "Both touch Qwen 3.5/3.6, but one is from_config text_config unwrapping and the other is CPU offload incompatibility in rotary embedding; not the same bug." - }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are load failures, but for different models and different symptoms: ScandiBERT model loading vs est-roberta tokenizer loading." + "reason": "Both are tokenizer/model loading failures, but they concern different models and likely different root causes; too broad to treat as the same bug." }, { "left": "issue:43540", "right": "issue:46858", "accept": false, - "reason": "Unrelated: Qwen3OmniMoe video input ValueError vs static-cache seq-length tensor arithmetic on non-CUDA backends." + "reason": "These cover unrelated failures: a Qwen3OmniMoe video-input ValueError versus a static-cache/torch.compile tensor arithmetic bug." }, { - "left": "issue:44077", - "right": "issue:44625", + "left": "issue:45357", + "right": "issue:46402", "accept": false, - "reason": "Distinct issues: patchtsmixer post_init policy vs Qwen3.5 num_labels propagation." + "reason": "Different model families and different failure modes: incorrect visual encoder keys on save versus missing text encoder weight loading." }, { - "left": "issue:45357", - "right": "issue:46402", + "left": "issue:44077", + "right": "issue:44625", "accept": false, - "reason": "Both concern model weights/config loading, but one is Qwen3.5 visual encoder keys on save_pretrained and the other is Sam3Model text encoder weights not loading." + "reason": "One is about an optional post_init policy in patchtsmixer; the other is config propagation of num_labels in Qwen3.5. Not the same code path." }, { "left": "issue:43644", "right": "issue:46665", "accept": false, - "reason": "Different bugs: non-persistent buffer junk in Transformers 5.0.0 vs inconsistent from_pretrained behavior inside Python classes." + "reason": "A non-persistent-buffer initialization bug and an inconsistent from_pretrained-in-Python-classes behavior are too different to merge." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Different subsystems and failures: loading physical-intelligence/fast on newer transformers vs PIL image processors wrongly requiring torchvision." + "reason": "Different underlying issues: a model loading failure for one repo versus PIL image processors wrongly requiring torchvision." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "AddedToken special-argument duplication is unrelated to ScandiBERT model loading." + "reason": "AddedToken keyword duplication and a model-loading failure for ScandiBERT are unrelated." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Different model-specific problems: Perceiver non-default resolution handling vs bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "One is a Perceiver image-size/interpolation issue; the other is a bfloat16 dtype mismatch in inference." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Gradient checkpointing in eval mode is a different issue from the DeepSpeed ZeRO-3 gradient growth regression." + "reason": "Both mention gradients, but one is about eval-mode checkpointing usage and the other about a ZeRO-3 training regression." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Related only at a broad classification level; one is config propagation, the other is a zero-loss bug with num_labels=1." - }, - { - "left": "issue:43526", - "right": "issue:43881", - "accept": false, - "reason": "Different bugs: BeitImageProcessorFast label reduction vs glm-4v-9b loading failure." + "reason": "These are separate classification-config bugs: num_labels propagation versus degenerate zero-loss behavior for num_labels=1." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Different models and causes: SigLIP2 doc/example errors and quantization failure vs OmDet-Turbo processor size mismatch." + "reason": "Both involve vision models, but one is a SigLIP2 doc/example problem and the other is an OmDet-Turbo processor size mismatch." }, { "left": "issue:41720", "right": "issue:45704", "accept": false, - "reason": "Completely different issues: Qwen3 auto device mapping cudaErrorAssert vs T5 using apex.FusedRMSNorm with a memory leak." + "reason": "Completely different problems: Qwen3 auto device mapping crash versus a T5/apex RMSNorm memory leak." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "A BeitImageProcessorFast label-reduction bug and a glm-4v-9b loading failure are unrelated." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "Both are Qwen3.5-related, but one is a tie_word_embeddings warning during LoRA fine-tuning and the other is wrong visual-encoder keys saved." + "reason": "A configuration warning about tie_word_embeddings is not the same bug as incorrect visual encoder keys on save_pretrained." }, { "left": "issue:43381", "right": "issue:45704", "accept": false, - "reason": "Different bugs in different models: eval-mode gradient checkpointing vs T5 RMSNorm memory leak." + "reason": "Different layers of the stack: eval-mode checkpointing restriction versus a T5 RMSNorm memory leak." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Different trainer failures: SFT input-embedding unpacking vs trainer.evaluate() breaking after trainer.train()." + "reason": "Both are trainer-related, but one is an embedding unpacking failure with TRL SFT and the other is evaluate-after-train behavior; no shared concrete bug." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "Both involve persistence/loading, but PEFT adapters on local checkpoints and sinq quantized save/load are different code paths and bugs." + "reason": "Local checkpoint loading with PEFT adapters and quantized save/load breakage are distinct persistence issues." }, { "left": "issue:46655", "right": "issue:46693", "accept": false, - "reason": "SmolVLM video padding bug and FA2 long-context performance regression are unrelated." + "reason": "Both are performance/processor-adjacent, but one is a video padding bug in SmolVLM and the other is a long-context FA2 generation regression." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Both mention tokens/config, but one is runtime post-processor updates and the other is custom token IDs lost on save/load." + "reason": "A tokenizer post-processor update bug and a DiaConfig custom-token-ID persistence bug are different state-management problems." }, { "left": "issue:45759", "right": "issue:45923", "accept": false, - "reason": "Different model/config issues: Qwen multimodal text_config unwrapping vs a missing supports_gradient_checkpointing flag on Nemotron-3-Nano-Omni." + "reason": "One is about unwrapping text_config in composite multimodal configs; the other is a missing supports_gradient_checkpointing flag." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "No shared underlying bug: T5 RMSNorm memory leak vs missing gradient_checkpointing support flag." + "reason": "A memory leak in apex FusedRMSNorm is unrelated to a missing capability flag on a trust_remote_code model." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "Different problems: PEFT local checkpoint loading vs quantization/offloading behavior." + "reason": "PEFT local checkpoint loading and quantization/offloading behavior are related only at a broad subsystem level, not the same bug." }, { "left": "issue:42886", "right": "issue:44843", "accept": true, - "reason": "Same underlying offline-mode tokenizer bug: both describe HF_HUB_OFFLINE causing tokenizer loading to fail, and 44843 identifies the specific unconditional model_info() call in _patch_mistral_regex." + "reason": "Both describe the same offline-mode failure path: tokenizer loading unconditionally triggers hub metadata access/model_info(), which breaks HF_HUB_OFFLINE." }, { "left": "issue:30064", "right": "issue:36010", "accept": false, - "reason": "Image segmentation-map processing failure and GenerationMixin import error are unrelated." + "reason": "Image processor void-segmentation-map handling and a GenerationMixin import error are unrelated." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "These are the same bug: identical Qwen2_5_VLProcessor.apply_chat_template crash on batched input when padding=False." + "reason": "These are the same crash report with effectively identical titles and behavior: Qwen2_5_VLProcessor.apply_chat_template fails on batched input when padding=False." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "Both are about SigLIP2, but one is a documentation/example mismatch and quantization failure while the other is a HF-vs-JAX implementation discrepancy." + "reason": "Both mention SigLIP2, but one is a documentation/example issue and the other is a implementation-vs-JAX discrepancy; not the same concrete fix." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "TokenizerBackend post-processor updates and BigBirdTokenizer special-token registration are distinct tokenizer bugs." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and symptoms: SigLIP2 nonsensical outputs versus CamemBERT masked LM prediction regression." + }, + { + "left": "issue:44464", + "right": "issue:47050", + "accept": false, + "reason": "Chunked generation with compiled forward and fp16 ONNX attention dtype mismatch are separate generation/export problems." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The two accepted soft pairs are well supported by the titles/body descriptions, and the rejected pairs are reasonably separated by bug, subsystem, or failure mode. No overstatement of duplicates or shared root cause stands out." + "feedback": "The summary is grounded in the packet: issue 46693 is a reasonable representative given its high discussion activity and broad performance-regression scope, and the two accepted soft pairs are conservative and well-supported by the titles. The remaining verdicts stay appropriately cautious and do not overstate sameness across clearly different bugs." }, "evaluator_used": true, "retried": false, @@ -5619,6 +5651,7 @@ "issue:39692", "issue:41720", "issue:43295", + "issue:43335", "issue:43421", "issue:43531", "issue:43540", @@ -5628,20 +5661,18 @@ "issue:43643", "issue:43653", "issue:43720", + "issue:43756", "issue:43782", "issue:43819", "issue:43873", "issue:43927", - "issue:43994", "issue:44062", "issue:44186", "issue:44220", "issue:44361", - "issue:44464", "issue:44488", "issue:44625", "issue:44661", - "issue:44671", "issue:44857", "issue:44898", "issue:44991", @@ -5664,9 +5695,6 @@ "issue:47050" ], "soft_pairs": [ - "issue:43421|issue:43653", - "issue:44464|issue:47050", - "issue:43994|issue:44671", "issue:46032|issue:46463", "issue:44857|issue:47050", "issue:43782|issue:45381", @@ -5680,256 +5708,239 @@ "issue:43295|issue:45834", "issue:43653|issue:43720", "issue:43550|issue:43927", - "issue:43596|issue:44220", "issue:43550|issue:44361", + "issue:43596|issue:44220", "issue:46829|issue:46832", - "issue:43531|issue:43643", "issue:44062|issue:44488", "issue:44625|issue:47050", + "issue:43531|issue:43643", "issue:43540|issue:46294", "issue:45381|issue:46693", "issue:45987|issue:46829", "issue:43611|issue:46665", "issue:30064|issue:36331", - "issue:44220|issue:44661", "issue:41720|issue:44991", + "issue:44220|issue:44661", "issue:43653|issue:43819", "issue:44186|issue:44898", "issue:44488|issue:47020", - "issue:43550|issue:43819" + "issue:43550|issue:43819", + "issue:43335|issue:43756" ], - "prepared_review_unit_hash": "2c313c8bcaa3ba2ce74600f8d4fe25abf9062a0721fe51565be59cc150b02cc4", + "prepared_review_unit_hash": "1e10afac0c2a64547170327c91b01238ad8020a77deeb73514342cd031634f5f", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37889, - "estimated_input_tokens": 9473, - "estimated_eval_tokens": 19202 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36768, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous and the only clear duplicate-like pair is the `get_seq_length()` contract bug. Most other soft pairs share broad keywords (tokenizers, quantization, video, loading) but point to different models, subsystems, or failure modes.", - "confidence": 0.87, - "canonical_issue_reason": "issue:45987 is the best canonical issue because it states the root API contract problem in broad form (`get_seq_length()` returning a tensor instead of an int), which also explains the more specific static-cache/compile failure in issue:46858.", + "summary": "This cluster is not a true duplicate set: it spans many unrelated bug reports across tokenizers, model loading, vision/video inputs, cache behavior, quantization, and configs. None of the soft pairs look like the same underlying code-path failure.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45987 is the strongest representative of the cluster\u2019s only clear duplicate pair and the most actionable root-cause report.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43421", - "right": "issue:43653", - "accept": false, - "reason": "Both concern tokenizer special-token handling, but one is runtime post-processor updates and the other is BigBird mask-token registration/decoding; not the same bug." - }, - { - "left": "issue:44464", - "right": "issue:47050", - "accept": false, - "reason": "Different code paths and failures: chunked generation with compiled forward vs ONNX fp16 attention dtype mismatch." - }, - { - "left": "issue:43994", - "right": "issue:44671", - "accept": false, - "reason": "Both are wrong-output reports, but on different models and different regressions; not a single underlying change." - }, { "left": "issue:46032", "right": "issue:46463", "accept": false, - "reason": "Unrelated issues: Mamba2 cache correctness vs T5 decoder input error-message behavior." + "reason": "Different subsystems and failures: Mamba2 cache correctness vs T5 decoder-input validation error messaging." }, { "left": "issue:44857", "right": "issue:47050", "accept": false, - "reason": "Both mention fp16/dtype, but the affected models and failure mechanisms are different." + "reason": "Unrelated models and bug types: AMP/CUDA loss crash vs ONNX fp16 attention dtype mismatch." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen VL/video related, but one is a weight-only load error and the other is wrong video position IDs; separate bugs." + "reason": "Both are Qwen vision-related, but they are different models and different failures (weight_only load error vs wrong video position IDs)." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Both touch `fix_mistral_regex`, but one is a tokenizer crash and the other is a codec-handling regression with a warning issue; not enough evidence of the same bug." + "reason": "Tokenizer regressions, but not the same bug: one is a backend_tokenizer attribute error, the other is codec handling/warning regression in a specific tokenizer version." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Tokenizer-special-token handling vs LayoutLMv2 NER/padding crashes; different concrete failures." + "reason": "Special-token post-processing behavior vs LayoutLMv2 NER/padding crash are different code paths." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Completely unrelated: tokenizer regression vs hub/revision forwarding in `list_repo_files`." + "reason": "Tokenizer regression vs HF Hub revision forwarding in CI; no shared underlying bug." }, { "left": "issue:44995", "right": "issue:46858", "accept": false, - "reason": "Both involve caches, but one is a stale indexer cache on second forward pass and the other is `get_seq_length()` returning a device tensor; different bugs." + "reason": "Both mention cache/indexing, but one is a stale model cache on second forward and the other is a static-cache sequence-length type issue under torch.compile." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "Both mention quantization, but one is a docs/example mismatch plus quantization failure and the other is offloading behavior with quantization." + "reason": "Documentation/example errors are unrelated to quantization offloading behavior." }, { "left": "issue:39692", "right": "issue:47050", "accept": false, - "reason": "Different models and different problems: SigLIP2 docs/quantization vs Canine fp16 ONNX dtype mismatch." + "reason": "Docs/example problems vs Canine ONNX export dtype mismatch are unrelated." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Packed-weight loading vs DAC latent conversion/STE mismatch; no shared underlying bug." + "reason": "Different models and failures: accelerate loading of packed weights vs DAC latent/STE mismatch." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Both are regressions, but one is processor/tokenizer API breakage and the other is long OCR index error in Kosmos2.5." + "reason": "Regression in processor/tokenizer access vs long OCR index error in Kosmos2.5; not the same defect." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Special-token decode issue vs BitNet packed-weight unpacking; unrelated subsystems." + "reason": "BigBird special-token registration bug is unrelated to BitNet packed-weight loading." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "Different failure modes: torch.compile/SDPA model issue vs config save/load losing custom token IDs." - }, - { - "left": "issue:43596", - "right": "issue:44220", - "accept": false, - "reason": "BertModel/DeepSpeed ZeRO-3 index error vs audio feature extraction issue; unrelated." + "reason": "torch.compile/SDPA crash in Bamba is unrelated to DiaConfig token ID persistence." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "One is a model/runtime compile failure, the other is a tokenizer attribute error; different bugs." + "reason": "Different model families and unrelated failures: compile/SDPA crash vs tokenizer AttributeError." }, { - "left": "issue:46829", - "right": "issue:46832", + "left": "issue:43596", + "right": "issue:44220", "accept": false, - "reason": "Different models and outputs: hidden_states misassignment vs RT-DETR feature-level forward TypeError." + "reason": "Deepspeed ZeRO-3/BertModel indexing vs audio feature extraction issue; no common bug." }, { - "left": "issue:43531", - "right": "issue:43643", + "left": "issue:46829", + "right": "issue:46832", "accept": false, - "reason": "Sliding-window Qwen3-MoE bug vs AutoConfig trust_remote_code field loss; unrelated." + "reason": "Both are output-shape/forward issues, but they affect different models and different invariants." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "Both are tokenizer-loading failures, but the titles point to different concrete issues and no shared code path is evident." + "reason": "AddedToken keyword conflict vs failure to load a specific tokenizer model are unrelated." }, { "left": "issue:44625", "right": "issue:47050", "accept": false, - "reason": "Qwen3.5 label propagation vs Canine fp16 ONNX dtype mismatch; unrelated." + "reason": "num_labels propagation in Qwen3.5 is unrelated to Canine fp16 ONNX export dtype mismatch." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to AutoConfig trust_remote_code field loss." }, { "left": "issue:43540", "right": "issue:46294", "accept": false, - "reason": "Both are video-input issues, but the affected models and failure mechanisms differ." + "reason": "Both are video-related Qwen issues, but the failing path differs: video input processing ValueError vs chat-template feature/token count mismatch." }, { "left": "issue:45381", "right": "issue:46693", "accept": false, - "reason": "Qwen2.5-VL video position IDs vs long-context FA2 performance regression; unrelated." + "reason": "Wrong video position IDs in Qwen2.5-VL is unrelated to a long-context FA2 performance regression." }, { "left": "issue:45987", "right": "issue:46829", "accept": false, - "reason": "Cache length contract bug vs VideoPrism output field misassignment; different components." + "reason": "StaticCache return type bug is unrelated to VideoPrism hidden_states/output packaging." }, { "left": "issue:43611", "right": "issue:46665", "accept": false, - "reason": "Both concern loading behavior, but one is about `base_model_prefix` and the other about `from_pretrained` inside Python classes; not the same issue." + "reason": "Both involve from_pretrained/loading, but one is base_model_prefix breakage in v5 and the other is inconsistent behavior inside Python classes." }, { "left": "issue:30064", "right": "issue:36331", "accept": false, - "reason": "Image segmentation processing vs a CustomTrainer keyword-argument regression; unrelated." + "reason": "Image segmentation map processing and CustomTrainer compute_loss signature mismatch are unrelated." }, { - "left": "issue:44220", - "right": "issue:44661", + "left": "issue:41720", + "right": "issue:44991", "accept": false, - "reason": "Audio feature extraction vs model-addition/template logic in tokenizer mapping; different areas." + "reason": "Qwen3 device-mapping CUDA assert is unrelated to tokenizer loading for EMBEDDIA/est-roberta." }, { - "left": "issue:41720", - "right": "issue:44991", + "left": "issue:44220", + "right": "issue:44661", "accept": false, - "reason": "Device-mapping CUDA assert vs tokenizer loading failure for a specific model; unrelated." + "reason": "Audio fbank extraction issue is unrelated to add-new-model-like failing inside TOKENIZER_MAPPING_NAMES." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "Tokenizer special-token registration vs DAC latent reconstruction mismatch; no overlap in underlying bug." + "reason": "Tokenizer special-token decode bug is unrelated to DAC.from_latents/forward mismatch." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer padding/NER crash vs Perceiver non-default resolution failure; different models and code paths." + "reason": "LayoutLMv2 tokenizer crashes and Perceiver non-default resolution failures are separate model-specific bugs." }, { "left": "issue:44488", "right": "issue:47020", "accept": false, - "reason": "Both are model/tokenizer loading issues, but the titles indicate different root causes and no clear common bug." + "reason": "Tokenizer loading failure for one model is unrelated to Unigram sequence-vs-dict construction in XLMRobertaTokenizer." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "torch.compile/SDPA model failure vs DAC missing-STE mismatch; unrelated." + "reason": "No shared code path: Bamba compile/SDPA crash vs DAC latent conversion bug." }, { - "left": "issue:45987", - "right": "issue:46858", - "accept": true, - "reason": "These look like the same root bug: `cache.get_seq_length()` / `StaticCache.get_seq_length()` returns a tensor instead of an int, and 46858 is the downstream static-cache/compile failure caused by that contract violation." + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "Both are model-configuration bugs, but they affect different architectures and different parameters." } ] }, @@ -5957,7 +5968,6 @@ "issue:36010", "issue:39692", "issue:43322", - "issue:43335", "issue:43421", "issue:43540", "issue:43550", @@ -5969,7 +5979,7 @@ "issue:43644", "issue:43697", "issue:43749", - "issue:43756", + "issue:43782", "issue:43818", "issue:44008", "issue:44186", @@ -5984,6 +5994,7 @@ "issue:44843", "issue:44877", "issue:44936", + "issue:44977", "issue:44991", "issue:45020", "issue:45072", @@ -6002,12 +6013,11 @@ "issue:47020" ], "soft_pairs": [ - "issue:43335|issue:43756", - "issue:43643|issue:44877", - "issue:45987|issue:46932", "issue:43618|issue:43697", - "issue:44008|issue:46858", + "issue:45987|issue:46932", + "issue:43643|issue:44877", "issue:44821|issue:45084", + "issue:44008|issue:46858", "issue:34634|issue:36010", "issue:44843|issue:45081", "issue:39692|issue:45834", @@ -6018,201 +6028,202 @@ "issue:44991|issue:47020", "issue:45183|issue:45479", "issue:44991|issue:46291", - "issue:43644|issue:43749", "issue:43322|issue:45685", + "issue:43644|issue:43749", "issue:43606|issue:45072", "issue:44561|issue:45020", - "issue:43550|issue:45072", "issue:43421|issue:44811", + "issue:43550|issue:45072", "issue:43577|issue:43818", - "issue:43582|issue:46512", "issue:44479|issue:45834", - "issue:44805|issue:44936" + "issue:44805|issue:44936", + "issue:43582|issue:46512", + "issue:43782|issue:44977" ], - "prepared_review_unit_hash": "aa460e2140f6b05fb463456a8f294348a24db2f480a8428461560c02a2a8bf6f", + "prepared_review_unit_hash": "4dde6af676c3b3634b0d323c1afb2746219805d953ba3b73fad3c9938251bab2", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35263, - "estimated_input_tokens": 8816, - "estimated_eval_tokens": 17888 + "serialized_chars": 35191, + "estimated_input_tokens": 8798, + "estimated_eval_tokens": 17852 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly a grab bag of unrelated model/tokenizer/config/regression reports rather than one duplicate cluster. I would reject all soft edges and, if a representative issue is needed, use the broad remote-code regression report as the umbrella reference.", - "confidence": 0.94, - "canonical_issue_reason": "Issue 45020 is the broadest umbrella report in the set ('Recent transformers versions break models using `remote_code`'), so it is the best representative anchor even though the surrounding issues are not true duplicates of it.", + "summary": "All soft pairs look like false positives: they share broad subsystems or regression themes, but not the same concrete bug/change. No clear canonical issue or PR emerges from this set.", + "confidence": 0.92, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45020 is the most suitable global issue because it is cross-model and regression-oriented, which makes it the best umbrella for related remote-code breakages.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43335", - "right": "issue:43756", - "accept": false, - "reason": "Different model families and different bugs: sparse-layer construction vs RoPE layer-dropping." - }, - { - "left": "issue:43643", - "right": "issue:44877", + "left": "issue:43618", + "right": "issue:43697", "accept": false, - "reason": "Both involve config loading, but one is missing fields with trust_remote_code and the other is strict config rejection for granite_speech." + "reason": "Both are Transformers v5 regressions, but they affect different models and different failures (CLIP attentions vs RTDetrV2 outputs)." }, { "left": "issue:45987", "right": "issue:46932", "accept": false, - "reason": "Unrelated: cache sequence-length type contract vs Gemma4 quantized embedder dtype casting." + "reason": "StaticCache seq-length semantics and Gemma4 quantized embedder dtype casting are unrelated code paths." }, { - "left": "issue:43618", - "right": "issue:43697", + "left": "issue:43643", + "right": "issue:44877", "accept": false, - "reason": "Both are v5 regressions, but they hit different models and different output-path problems." + "reason": "Both concern config loading, but one is missing fields with trust_remote_code and the other is strict granite_speech config validation; not the same bug." }, { - "left": "issue:44008", - "right": "issue:46858", + "left": "issue:44821", + "right": "issue:45084", "accept": false, - "reason": "Different subsystems: Gemma3n forward variable collision vs static cache sequence-length device-tensor bug." + "reason": "AutoImageProcessor URL loading and template-node compilation errors are unrelated." }, { - "left": "issue:44821", - "right": "issue:45084", + "left": "issue:44008", + "right": "issue:46858", "accept": false, - "reason": "No shared underlying bug; image processor URL loading is unrelated to template-node compilation failure." + "reason": "Gemma3n variable-name collision in forward() and cache.get_seq_length() returning a device tensor are different bugs." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "Different areas entirely: Bark voice_preset handling vs a GenerationMixin import error." + "reason": "BarkProcessor voice_preset handling and the GenerationMixin import error are unrelated." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Same helper function name, but different failures: unconditional hub lookup offline vs missing backend_tokenizer attribute." + "reason": "Both hit _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer attribute crash; separate failure modes." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "Unrelated documentation/example errors vs a Kosmos2.5 long-OCR index error." + "reason": "SigLIP2 doc-example/model mismatch and Kosmos2.5 OCR long-input indexing are unrelated." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "Different tokenizer problems for different models; not the same code-path bug." + "reason": "LayoutLMv2Tokenizer NER/padding crashes and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are different tokenizer issues." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Different models and different failure modes: video input validation vs Gemma3n attribute collision." + "reason": "Qwen3OmniMoe video-processing ValueError and Gemma3n audio mask collision are unrelated multimodal bugs." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Both are v5 output mismatches, but for different models and likely different causes." + "reason": "Both are v5 output regressions, but they concern different models and different behavior changes." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty transcription handling is unrelated to Kosmos2.5 OCR indexing." + "reason": "Whisper empty-transcription generation and Kosmos2.5 long OCR input indexing are not the same problem." }, { "left": "issue:44991", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer-loading regressions, but the affected models and failure mechanisms differ." + "reason": "Tokenizer loading for EMBEDDIA/est-roberta and XLMRobertaTokenizer Unigram init are different root causes." }, { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Input validation messaging and degenerate classification loss are unrelated bugs." + "reason": "Input validation errors and single-label classification zero-loss behavior are separate sequence-classification issues." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer loading for a specific model is unrelated to pipelines import failing on kernels metadata requirements." + "reason": "Tokenizer loading failure and pipelines import failing on kernels v0.15 are unrelated." }, { - "left": "issue:43644", - "right": "issue:43749", + "left": "issue:43322", + "right": "issue:45685", "accept": false, - "reason": "Non-persistent buffer initialization and FSDP CPU RAM-efficient loading are different issues." + "reason": "Llava Next loading segfault and an MPS histogram dtype limitation are unrelated." }, { - "left": "issue:43322", - "right": "issue:45685", + "left": "issue:43644", + "right": "issue:43749", "accept": false, - "reason": "Different models and different runtime errors: Llava Next segfault vs MOE MPS histogram limitation." + "reason": "Non-persistent buffer junk and FSDP CPU RAM efficient loading are different breakages." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Both mention device/dtype mismatch, but they involve different models and different execution paths." + "reason": "CPU offload device mismatch for Bark and bfloat16 dtype mismatches for SwitchTransformers/TimmWrapperModel are different concrete issues." }, { "left": "issue:44561", "right": "issue:45020", "accept": false, - "reason": "Related theme of remote_code regressions, but one is a specific removed helper and the other is a broader umbrella report." - }, - { - "left": "issue:43550", - "right": "issue:45072", - "accept": false, - "reason": "Different bugs in different models: SDPA torch.compile failure vs bfloat16 dtype mismatch." + "reason": "Both mention remote_code breakage, but 45020 is a broad umbrella and does not identify the same specific code-path as the is_torch_fx_available removal." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Both touch token handling, but one is runtime post-processor updates and the other is skip_special_tokens behavior in Whisper decoding." + "reason": "Runtime special-token/post-processor updates and Whisper skip_special_tokens decoding are unrelated tokenizer/processor behaviors." }, { - "left": "issue:43577", - "right": "issue:43818", + "left": "issue:43550", + "right": "issue:45072", "accept": false, - "reason": "Different architectures and unrelated problems: BLIP2 dtype propagation vs Video-LLaVA temporal-attention/weight-sharing bug." + "reason": "torch.compile+SDPA failure in Bamba-9B-v2 and dtype mismatches in other models are different problems." }, { - "left": "issue:43582", - "right": "issue:46512", + "left": "issue:43577", + "right": "issue:43818", "accept": false, - "reason": "Both involve dtype on Apple/creation paths, but one is a TypeError in warmup and the other is stale config dtype after from_config." + "reason": "BLIP2 dtype propagation and Video-LLaVA missing temporal attention/shared weights are unrelated multimodal issues." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Different multimodal regressions: Qwen video input issue vs Kosmos2.5 OCR indexing." + "reason": "Qwen video-input regression and Kosmos2.5 OCR long-input indexing are different model failures." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Index/mask shape mismatch is unrelated to trainer.evaluate() failing after training." + "reason": "Mask shape mismatch IndexError and trainer.evaluate() failing after train() do not point to the same bug." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon warmup TypeError and stale config.dtype after from_config(dtype=...) are unrelated dtype/runtime issues." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation issues are different code paths." } ] }, @@ -6250,6 +6261,7 @@ "issue:43653", "issue:43716", "issue:43720", + "issue:43742", "issue:43749", "issue:43782", "issue:43819", @@ -6264,8 +6276,8 @@ "issue:44671", "issue:44857", "issue:44912", - "issue:44977", "issue:45003", + "issue:45020", "issue:45042", "issue:45072", "issue:45081", @@ -6285,226 +6297,223 @@ "issue:47217" ], "soft_pairs": [ - "issue:43782|issue:44977", "issue:43825|issue:43994", "issue:43388|issue:46032", "issue:46294|issue:46463", - "issue:43716|issue:44008", "issue:43931|issue:46846", + "issue:43716|issue:44008", "issue:28218|issue:29159", "issue:43540|issue:45381", "issue:44610|issue:45072", - "issue:43749|issue:44291", "issue:43720|issue:45072", + "issue:43749|issue:44291", "issue:43645|issue:46665", "issue:34634|issue:36331", "issue:43825|issue:43827", "issue:42175|issue:45081", "issue:44671|issue:45701", "issue:43577|issue:43582", - "issue:46175|issue:46932", "issue:45003|issue:46285", - "issue:43582|issue:44857", - "issue:43653|issue:44912", + "issue:46175|issue:46932", "issue:43262|issue:47217", "issue:46697|issue:46728", - "issue:43782|issue:45325", + "issue:43582|issue:44857", + "issue:43653|issue:44912", "issue:43262|issue:43819", - "issue:44589|issue:45042" + "issue:43782|issue:45325", + "issue:44589|issue:45042", + "issue:43742|issue:45020" ], - "prepared_review_unit_hash": "55b0230ae54b77101a9c2d628b92a46af6d2edf0b1c4606e8694f45569044468", + "prepared_review_unit_hash": "c176d9dcfecaa6d9f15b4a460a5d7f99e35ff84ac0afb785dfc37e820df22a0b", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 35387, - "estimated_input_tokens": 8847, - "estimated_eval_tokens": 17950 + "serialized_chars": 35943, + "estimated_input_tokens": 8986, + "estimated_eval_tokens": 18228 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous and issue-only. The only near-match is the tokenizer pair 28218/29159, but the evidence is limited to title-level similarity, so it should be treated as related/possibly duplicate rather than confirmed duplicate.", - "confidence": 0.21, - "canonical_issue_reason": "issue:29159 is the more general tokenizer-consistency report of the only close pair, but the packet only supports a cautious related/possibly-duplicate relationship with issue:28218, not a strong subsumption claim.", + "summary": "The cluster is largely heterogeneous: the soft pairs are mostly superficial similarities across different models, tokenizers, loaders, or processors rather than the same underlying bug/change. I would not merge any of the proposed pairs.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue emerges; the items span many unrelated bugs and only share broad transformer-like themes.", "canonical_pr_reason": null, - "best_issue_reason": "issue:29159 is the best single representative of the only close tokenizer pair because it states the broader inconsistency framing, while the rest of the cluster is unrelated.", + "best_issue_reason": "No global best issue for deduplication; the cluster does not center on one underlying defect.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43782", - "right": "issue:44977", - "accept": false, - "reason": "Different Qwen3 problems: weight-only loading failure vs flash-attention generation behavior." - }, { "left": "issue:43825", "right": "issue:43994", "accept": false, - "reason": "Runtime error-message wording/removal issue vs a model output-quality bug; not the same code path." + "reason": "Different problems: a pipeline error-message regression vs. a model producing bad outputs. Same broad area, not the same bug." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "Metrics label dropping and Mamba cache correctness are unrelated bugs." + "reason": "Unrelated subsystems: metric gathering vs. Mamba2 cache behavior." }, { "left": "issue:46294", "right": "issue:46463", "accept": false, - "reason": "Qwen3.5 video chat-template mismatch vs T5 decoder-input validation message; different subsystems." + "reason": "Different validation paths and models; one is Qwen video chat templating, the other is T5 decoder-input error messaging." }, { - "left": "issue:43716", - "right": "issue:44008", + "left": "issue:43931", + "right": "issue:46846", "accept": false, - "reason": "Different multimodal loading/typing bug vs a Gemma3n forward variable-collision bug." + "reason": "Different Qwen model-loading/shape issues with no clear shared code path." }, { - "left": "issue:43931", - "right": "issue:46846", + "left": "issue:43716", + "right": "issue:44008", "accept": false, - "reason": "Both involve Qwen3/VL, but one is a weight-shape load error and the other is a TP-plan omission." + "reason": "Different multimodal components and failure modes: image-preprocessor dtype mismatch vs. Gemma3n forward variable collision." }, { "left": "issue:28218", "right": "issue:29159", "accept": false, - "reason": "They are plausibly related tokenizer reports, but the available evidence is only title-level similarity, not enough to confirm the same underlying bug." + "reason": "Only loosely related tokenizer behavior; one is added-token spacing, the other is a broad slow/fast inconsistency report." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Both are video-input bugs, but they affect different models and likely different preprocessing paths." + "reason": "Different models and different video-position/processing bugs." }, { "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "Image-size preprocessing mismatch and bfloat16 dtype mismatch are unrelated." + "reason": "Processor-size mismatch and dtype mismatch are unrelated." }, { - "left": "issue:43749", - "right": "issue:44291", + "left": "issue:43720", + "right": "issue:45072", "accept": false, - "reason": "Different loading mechanisms: FSDP CPU RAM-efficient loading vs init_empty_weights argument handling." + "reason": "Packed-weight loading bug is unrelated to bfloat16 dtype mismatch." }, { - "left": "issue:43720", - "right": "issue:45072", + "left": "issue:43749", + "right": "issue:44291", "accept": false, - "reason": "BitNet packed-weight loading failure and a dtype-mismatch inference bug are not the same issue." + "reason": "Both involve loading, but one is FSDP CPU RAM-efficient loading and the other is init_empty_weights argument compatibility." }, { "left": "issue:43645", "right": "issue:46665", "accept": false, - "reason": "Notebook custom-model initialization breakage and from_pretrained-in-class inconsistency are different behaviors." + "reason": "Different initialization contexts and failure modes; not the same defect." }, { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss are unrelated." + "reason": "Completely different areas: BarkProcessor preset handling vs. trainer loss signature changes." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "One is a runtime error-message issue; the other is stale documentation. Related topic, not the same bug." + "reason": "One is an error message bug, the other is stale documentation referencing pipeline(); related topic, not one fix." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Package/backend dependency issue vs tokenizer regex patch crash; different code paths." + "reason": "TensorFlow backend packaging and tokenizer regex patching are unrelated." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "CamemBERT MLM regression is not the same as the broad tokenization-change report." + "reason": "Tokenization-version changes may be adjacent, but these describe different concrete failures and are not clearly the same bug." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "BLIP2 dtype propagation and Apple Silicon warmup TypeError are unrelated." + "reason": "Different dtype/load state issue vs. Apple Silicon allocator TypeError." }, { - "left": "issue:46175", - "right": "issue:46932", + "left": "issue:45003", + "right": "issue:46285", "accept": false, - "reason": "Both touch Gemma4 embedders, but one is a huge one-hot allocation and the other is dtype casting for quantized embedders." + "reason": "sys.modules access safety and CLIP model structure inconsistency are unrelated." }, { - "left": "issue:45003", - "right": "issue:46285", + "left": "issue:46175", + "right": "issue:46932", "accept": false, - "reason": "sys.modules access safety and CLIP text-model structure changes are unrelated." + "reason": "One is a huge one-hot position embedding memory issue; the other is dtype casting for quantized embedders." }, { - "left": "issue:43582", - "right": "issue:44857", + "left": "issue:43262", + "right": "issue:47217", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError and CUDA AMP crash in a loss module are unrelated." + "reason": "Audio chat-template sampling-rate default and VL image-placeholder handling are different bugs." }, { - "left": "issue:43653", - "right": "issue:44912", + "left": "issue:46697", + "right": "issue:46728", "accept": false, - "reason": "BigBird special-token registration and MXFP4 load fallback are unrelated." + "reason": "Both concern image processing, but the underlying logic differs: over-padding vs. patch-count computation." }, { - "left": "issue:43262", - "right": "issue:47217", + "left": "issue:43582", + "right": "issue:44857", "accept": false, - "reason": "Audio chat-template sampling-rate defaults and VL image-placeholder handling are different multimodal bugs." + "reason": "Apple Silicon allocator warmup TypeError is unrelated to AMP/CUDA loss crashes." }, { - "left": "issue:46697", - "right": "issue:46728", + "left": "issue:43653", + "right": "issue:44912", "accept": false, - "reason": "Both are image-processing sizing bugs, but they concern different models and different calculations." + "reason": "Tokenizer special-token registration and MXFP4 weight loading are unrelated." }, { - "left": "issue:43782", - "right": "issue:45325", + "left": "issue:43262", + "right": "issue:43819", "accept": false, - "reason": "Qwen3VL weight-only load failure and Qwen2.5-VL rope-index scaling are unrelated." + "reason": "Audio template defaults and DAC forward/from_latents mismatch are unrelated." }, { - "left": "issue:43262", - "right": "issue:43819", + "left": "issue:43782", + "right": "issue:45325", "accept": false, - "reason": "Audio processor chat-template defaults and DAC latent decoding behavior are unrelated." + "reason": "Different Qwen-VL bugs: weight_only loading error vs. incorrect temporal position scaling." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup failure and PIL backend torchvision dependency handling are unrelated." + "reason": "Storage-object lookup failure and torchvision dependency regression are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "KeyError loading one model is not the same as remote_code compatibility breaking across versions." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The cluster looks mostly issue-only and heterogeneous, and the 28218/29159 pair is a plausible duplicate candidate. However, the summary overstates the evidence by asserting that issue:29159 is the broadest tokenizer regression and that it plausibly subsumes issue:28218; the packet only supports a conservative 'related/possibly duplicate' read from the titles and soft similarity. The canonical issue reasoning is therefore too strong for the available evidence." - }, - "evaluator_used": true, - "retried": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -6549,6 +6558,7 @@ "issue:43927", "issue:44060", "issue:44368", + "issue:44479", "issue:44554", "issue:44610", "issue:44811", @@ -6556,7 +6566,6 @@ "issue:44936", "issue:44977", "issue:44991", - "issue:45020", "issue:45042", "issue:45081", "issue:45084", @@ -6572,7 +6581,6 @@ "issue:47050" ], "soft_pairs": [ - "issue:43742|issue:45020", "issue:45372|issue:46326", "issue:43550|issue:44912", "issue:36010|issue:43645", @@ -6580,8 +6588,8 @@ "issue:43257|issue:43329", "issue:43531|issue:45042", "issue:43526|issue:43697", - "issue:42175|issue:46291", "issue:43638|issue:44936", + "issue:42175|issue:46291", "issue:43452|issue:44991", "issue:43906|issue:46209", "issue:43819|issue:43927", @@ -6591,212 +6599,213 @@ "issue:43329|issue:45600", "issue:43526|issue:44554", "issue:42490|issue:43257", - "issue:43643|issue:45759", "issue:43746|issue:44811", "issue:43854|issue:44977", + "issue:43643|issue:45759", "issue:45859|issue:47050", "issue:38175|issue:44610", "issue:39692|issue:44912", "issue:43262|issue:44811", "issue:43257|issue:43696", + "issue:43329|issue:44479", "issue:42915|issue:43742" ], - "prepared_review_unit_hash": "aac201a1a4b7df40c4ecad038fbe929c041a4ad7896a8062a2c7e2822e6927b4", + "prepared_review_unit_hash": "498c64da19d01704364425e7e94036ba32d5419ebd0e1c318b55daec7a9bfcc2", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36510, - "estimated_input_tokens": 9128, - "estimated_eval_tokens": 18512 + "serialized_chars": 36549, + "estimated_input_tokens": 9138, + "estimated_eval_tokens": 18532 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous: most items are distinct model-, processor-, backend-, or training-specific bugs rather than one duplicate family. No soft edge pair looks like the same underlying defect, so all candidate links should be rejected.", - "confidence": 0.94, - "canonical_issue_reason": "issue:45020 is the broadest and most umbrella-like issue here, centered on a recent-version regression affecting remote_code loading, with enough activity to serve as the cluster\u2019s representative issue.", + "summary": "This cluster is highly heterogeneous: the soft-similarity pairs are mostly superficial (shared buzzwords like loading, tokenizer, MoE, processor, or model naming) but do not describe the same underlying bug/change. No soft edge should be merged.", + "confidence": 0.95, + "canonical_issue_reason": "No true duplicate center emerges from this set; if one issue must anchor the cluster, issue 43257 is the most substantive and well-scoped among the Qwen/MoE-loading reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the most generally useful issue to keep as the representative of this mixed cluster because it describes a cross-cutting Transformers regression rather than a single narrow model symptom.", + "best_issue_reason": "Issue 43257 is the clearest, most concrete bug report in the set, but it is only a local anchor, not a universal representative of the whole heterogeneous cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43742", - "right": "issue:45020", - "accept": false, - "reason": "Different failures: a MobileLLM load-time KeyError vs a broad remote_code regression. Not the same concrete bug." - }, { "left": "issue:45372", "right": "issue:46326", "accept": false, - "reason": "Different subsystems and symptoms: Gemma 4 processor import failure vs mistral_common continue_final_message behavior." + "reason": "Both involve mistral_common-related tokenizers, but one is an import error in Gemma 4 processor loading and the other is a continue_final_message behavior bug; different failure modes and code paths." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "Different model paths and causes: torch.compile+SDPA on Bamba vs MXFP4 quantization fallback on git-oss-20b." + "reason": "Both mention Qwen/OSS models and quantization/attention, but one is a torch.compile + SDPA failure and the other is an MXFP4 loading fallback; not the same bug." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "GenerationMixin import error is unrelated to custom model initialization in Jupyter notebooks." + "reason": "GenerationMixin import error vs Jupyter custom-model initialization breakage in Transformers 5.0; unrelated symptoms and fixes." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "One is an import-path regression, the other is a custom Trainer API mismatch. Not the same bug." + "reason": "Importing GenerationMixin and CustomTrainer.compute_loss signature changes are unrelated API breakages." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to the multimodal token-counting video-branch NameError." + "reason": "Both are Qwen3 MoE-related, but one is about accelerate+deepspeed weight conversion and the other is a multimodal token-counting video-branch NameError/undefined variable bug." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Sliding-window behavior on Qwen3-MoE and PIL backend torchvision dependency handling are different issues." + "reason": "Both concern processor/input handling, but one is a Qwen3-MoE sliding_window issue and the other is a PIL backend dependency regression on torchvision." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to RTDetrV2 output drift in v5." + "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output drift are different model-specific bugs with different code paths." }, { - "left": "issue:42175", - "right": "issue:46291", + "left": "issue:43638", + "right": "issue:44936", "accept": false, - "reason": "Torch extra missing tensorflow backend is a packaging issue; kernels v0.15 import failure is a separate dependency requirement problem." + "reason": "Deepspeed ZeRO-3 index error during training and trainer.evaluate() failing after train() are distinct trainer/runtime issues." }, { - "left": "issue:43638", - "right": "issue:44936", + "left": "issue:42175", + "right": "issue:46291", "accept": false, - "reason": "DeepSpeed ZeRO-3 indexing failure during training is not the same as evaluate() failing after train()." + "reason": "Torch backend installation vs kernels v0.15 import/version constraint is a dependency-management mismatch, not the same issue." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "gguf_file breaks tokenizer/model loading APIs broadly, while est-roberta is a specific tokenizer regression." + "reason": "Both are tokenizer/model loading problems, but gguf_file support failure and a specific tokenizer load regression for EMBEDDIA/est-roberta are not the same bug." }, { "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 earlier issue and an FP8 inference-on-B200 bug are unrelated." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "DAC latent reconstruction mismatch and DiaConfig save/load token-ID loss are different model/config bugs." + "reason": "DAC latent mismatch and DiaConfig custom-token-ID loss are different model/config correctness issues." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Both are attribute errors, but they occur in different code paths and models: Mistral regex patching vs Gemma4 processor loading." + "reason": "Both hit Gemma/Mistral tokenizer-related loading paths, but one is a Mistral regex patch crash and the other is a Gemma4Processor missing _tokenizer in transformers serve." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both mention warnings around tied weights, but one is an incorrect internal tie and the other is a user-facing warning during LoRA fine-tuning." + "reason": "Both involve Qwen3.x MoE tied-weight warnings, but one is a specific incorrect weight-tying warning and the other is a LoRA tuning warning about tie_word_embeddings; not the same defect." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template-node compilation error is unrelated to CLIPTokenizer's model_max_length setting." + "reason": "Template compilation error and CLIPTokenizer max_length constant are unrelated tokenizer/template issues." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Video multimodal token counting bug is unrelated to stale auto_mappings references breaking CI." + "reason": "Video token counting bug and auto_mappings.py CI break from removed configs are unrelated." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "Image label reduction in BeitImageProcessorFast is not the same as an MPS attention correctness issue." + "reason": "Beit label reduction bug and an MPS attention correctness issue have no shared code path." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Trailing-slash local path warning and Qwen3 MOE weight conversion under accelerate/deepspeed are unrelated load-time behaviors." - }, - { - "left": "issue:43643", - "right": "issue:45759", - "accept": false, - "reason": "Missing fields from AutoConfig with trust_remote_code and unwrapping text_config in composite multimodal configs are different configuration-path bugs." + "reason": "Loading a local model path with a trailing slash and Qwen3 MoE weight conversion under accelerate+deepspeed are unrelated loader behaviors." }, { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "PEFT local-checkpoint loading and Whisper batch_decode skip_special_tokens handling are unrelated." + "reason": "PEFT adapter local checkpoint loading and Whisper processor.batch_decode skipping special tokens are different processor/model bugs." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "GLM-4.7-Flash unit-test load failure is not the same as Qwen3.5 flash-attention generation issues." + "reason": "Unit-test model loading for zai-org/GLM-4.7-Flash and Qwen3.5 flash-attention generation failure are separate model-specific issues." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "trust_remote_code AutoConfig field loss and AutoModelForCausalLM.from_config not unwrapping text_config are distinct config-loading bugs." }, { "left": "issue:45859", "right": "issue:47050", "accept": false, - "reason": "CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding is unrelated to CanineModel ONNX fp16 dtype mismatch." + "reason": "CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding and Canine fp16 ONNX export dtype mismatch are unrelated." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and OmDet-Turbo 640px/224px preprocessing mismatch are different vision bugs." + "reason": "SigLIP2 zero probabilities and OmDet-Turbo processor/model input-size mismatch are different vision-model issues." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 docs example errors are not the same concrete bug as git-oss-20b MXFP4 quantization fallback." + "reason": "SigLIP2 documentation example errors and git-oss-20b MXFP4 loading fallback are unrelated; one is docs/example correctness, the other quantization behavior." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio apply_chat_template sampling-rate defaulting and Whisper batch_decode token skipping are distinct processor behaviors." + "reason": "Audio apply_chat_template default sample-rate bug and Whisper batch_decode skip_special_tokens bug affect different processor methods." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MOE weight conversion failure is unrelated to GPT-oss-20b CUDA OOM." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and GPT-oss-20b CUDA OOM are not the same failure." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Undefined multimodal token helper variables in the video branch and a broader v5.3.0 video-input regression across multiple VLMs are related only by domain, not by concrete bug." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and MobileLLM load KeyError are different model-specific issues." + "reason": "FineGrainedFP8Config failure for Qwen3Moe and KeyError loading facebook/MobileLLM-125M are separate loader/config issues." } ] }, @@ -6846,12 +6855,12 @@ "issue:43931", "issue:44291", "issue:44387", - "issue:44479", "issue:44554", "issue:44568", "issue:44617", "issue:44756", "issue:44821", + "issue:44857", "issue:44912", "issue:44987", "issue:45042", @@ -6866,220 +6875,221 @@ "issue:46421", "issue:46424", "issue:46701", + "issue:46832", "issue:47145" ], "soft_pairs": [ - "issue:43329|issue:44479", "issue:43257|issue:43931", "issue:43329|issue:43450", "issue:44291|issue:45042", - "issue:44912|issue:47145", "issue:34567|issue:40990", + "issue:44912|issue:47145", "issue:42915|issue:44821", "issue:44387|issue:44617", "issue:29159|issue:34567", "issue:45459|issue:45803", "issue:46421|issue:46424", "issue:43525|issue:46291", - "issue:28282|issue:29870", "issue:43650|issue:43906", + "issue:28282|issue:29870", "issue:44987|issue:45538", "issue:43425|issue:46291", "issue:43749|issue:44756", "issue:28282|issue:34689", "issue:43540|issue:46701", - "issue:43454|issue:43577", "issue:44568|issue:45356", - "issue:42175|issue:43824", + "issue:43454|issue:43577", "issue:33357|issue:34567", + "issue:42175|issue:43824", "issue:39379|issue:44554", "issue:43697|issue:45357", - "issue:45137|issue:45405" + "issue:45137|issue:45405", + "issue:44857|issue:46832" ], - "prepared_review_unit_hash": "8a7c1e3bd71ecda3349a4816c5dcf6c2a1c19d75d2b3fc4b61c88e1aecf92ed6", + "prepared_review_unit_hash": "a33381f07fee0f46fc75d93cf74aa210c67e46d28c6377c1053a3efce52b3ab7", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 34975, - "estimated_input_tokens": 8744, - "estimated_eval_tokens": 17744 + "serialized_chars": 35518, + "estimated_input_tokens": 8880, + "estimated_eval_tokens": 18016 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft duplicate edges should be merged here: the pairs mostly share a broad subsystem (tokenizers, video, quantization, loading, memory) but describe different concrete bugs and code paths. The strongest central issue in this set is the FSDP CPU RAM efficient loading regression.", - "confidence": 0.68, - "canonical_issue_reason": "issue:43749 is the best cluster anchor: it has the most discussion and inbound references, and it describes a broad, clearly impactful core-loading regression rather than a narrow model-specific edge case.", + "summary": "This cluster is highly heterogeneous: the soft pairs mostly share broad keywords (tokenizer, video, quantization, cache, loading) but describe different failures and code paths, so none are safe duplicate merges.", + "confidence": 0.99, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43749 is the best overall issue to keep as the representative problem because it is broadly applicable, well-discussed, and likely to attract/organize related reports.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43329", - "right": "issue:44479", - "accept": false, - "reason": "Both involve video/multimodal processing, but 43329 is an undefined-variable bug in multimodal token counting while 44479 is a broader video-input regression across several model families." - }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Different failures in different Qwen families: one is MOE weight conversion with accelerate/deepspeed, the other is a VL weight-shape mismatch on model load." + "reason": "Both mention Qwen3 model loading, but one is about MoE weight conversion with accelerate/deepspeed and the other is a Qwen3-VL shape-mismatch load error." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "One is an undefined helper/merge_size issue in multimodal token counting; the other is batched video processor shape handling. Related area, not the same bug." + "reason": "Both are video-related, but one is an undefined-variable bug in multimodal token counting and the other is a batched video-processor shape bug." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "Both are loading-time issues, but one is an init_empty_weights argument mismatch and the other is an image-processor torchvision dependency problem." + "reason": "Different code paths: init_empty_weights argument handling versus PIL backend processors wrongly requiring torchvision." }, { - "left": "issue:44912", - "right": "issue:47145", + "left": "issue:34567", + "right": "issue:40990", "accept": false, - "reason": "Both touch quantization, but MXFP4 falling back to bf16 and FP8 frozen checkpoints ignoring weight_scale are distinct mechanisms and fixes." + "reason": "TrainerState token counting and high perplexity on a model are unrelated symptoms and subsystems." }, { - "left": "issue:34567", - "right": "issue:40990", + "left": "issue:44912", + "right": "issue:47145", "accept": false, - "reason": "Trainer token-count bookkeeping and a perplexity regression on gpt-oss-20b are unrelated problems." + "reason": "Both involve quantized checkpoints, but one is MXFP4 falling back to bf16 and the other is FP8 weight_scale not being applied; different formats and failures." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe and AutoImageProcessor URL loading are unrelated code paths." + "reason": "Qwen3Moe FP8 config failure is unrelated to loading AutoImageProcessor from a URL." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Both can manifest as OOM, but one is increased reserved memory under int4 quantization and the other is a SAM3Video model memory footprint issue." + "reason": "Both are OOM reports, but they come from different models and causes: int4 reserved-memory growth versus Sam3Video memory use." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is not the same as TrainerState token counting not updating." + "reason": "Tokenizer slow/fast inconsistency is a different bug from a TrainerState counter not updating." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Both mention exception handling masking errors, but they are different modules with different triggering conditions and fixes." + "reason": "Both involve exception handling, but they are different helpers and different swallowed-error behavior." }, { "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 being ignored except for Quantized Cache." + "reason": "Both touch cache logic, but one is lazy-init recompilation with chunked prefill and the other is max_cache_length being ignored outside quantized cache." }, { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "Missing pad_token_id on Llama4Config and kernels-v0.15 pipeline import failure are unrelated." + "reason": "Llama4Config missing pad_token_id is unrelated to kernels import/version requirements in pipelines." }, { - "left": "issue:28282", - "right": "issue:29870", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "PyTorch import-missing error and Bart generate decoder_attention_mask unbound-local error are unrelated." + "reason": "No substantive duplicate signal: one is a vague placeholder issue and the other is a reproduction of a different upstream bug." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:28282", + "right": "issue:29870", "accept": false, - "reason": "No meaningful overlap: a vague data request versus an isolated reproduction issue." + "reason": "PyTorch missing ImportError and Bart generate decoder_attention_mask UnboundLocalError are unrelated." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "A model-loading regression for physical-intelligence/fast and CLIPTokenizer max_length behavior are separate tokenizer/model issues." + "reason": "Different issues: a load regression for physical-intelligence/fast versus CLIPTokenizer's model_max_length constant." }, { "left": "issue:43425", "right": "issue:46291", "accept": false, - "reason": "Torch 2.10 incompatibility and kernels v0.15 pipeline import requirements are different dependency failures." + "reason": "Torch 2.10 compatibility and kernels import/version enforcement are different dependency problems." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "Both are memory/loading themed, but FSDP CPU RAM efficient loading and disabling mmap on Strix Halo are distinct issues." + "reason": "FSDP CPU RAM efficient loading and disabling mmap on Strix Halo are not the same bug or code path." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Missing PyTorch import and Llama 3.2 Vision model loading breakage are unrelated." + "reason": "Missing PyTorch import and Llama 3.2 90B Vision model-loading regression are unrelated." }, { "left": "issue:43540", "right": "issue:46701", "accept": false, - "reason": "Different multimodal regressions affecting different models and APIs; not the same underlying bug." - }, - { - "left": "issue:43454", - "right": "issue:43577", - "accept": false, - "reason": "AyaVision lm_head weight tying and Blip2 dtype propagation are separate model-loading problems." + "reason": "Qwen3OmniMoe video-input ValueError and DiffusionGemma pipeline failure are different model/pipeline bugs." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different tokenizers and different failure modes." + "reason": "Different tokenizer regressions: mdeberta add_special_tokens behavior versus Kimi-K2.5 codec handling/warning issues." }, { - "left": "issue:42175", - "right": "issue:43824", + "left": "issue:43454", + "right": "issue:43577", "accept": false, - "reason": "Backend install dependency packaging and a missing exported model class are unrelated." + "reason": "AyaVision weight tying causing bad generations is unrelated to BLIP2 dtype remaining float32." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "A MacOS bus error in CLIP model loading and TrainerState token counting are unrelated." + "reason": "MacOS bus error in CLIP loading and a TrainerState token counter bug do not overlap." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Missing tensorflow backend in a torch install is unrelated to importing a missing Qwen2_5_VL class." }, { "left": "issue:39379", "right": "issue:44554", "accept": false, - "reason": "Both are correctness issues, but one is an mllama multimodal mask bug and the other is an MPS attention-dimension correctness issue." + "reason": "Both are correctness bugs, but one is an mllama mask logic issue and the other is an MPS attention backend issue with different head dimensions." }, { "left": "issue:43697", "right": "issue:45357", "accept": false, - "reason": "RTDetrV2 output regression and Qwen3.5 save_pretrained key handling are different regression types in different paths." + "reason": "RT-DETR output changes and Qwen3.5 save_pretrained visual-key corruption are separate regressions." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque pop failure and a PEFT minimum-version bump are unrelated." + "reason": "DeepSpeed ZeRO3 deque failure and a PEFT version bump are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:46832", + "accept": false, + "reason": "Different RT-DETR/LwDetr bugs with different failure modes: AMP crash versus feature-level mismatch TypeError." } ] }, @@ -7114,6 +7124,7 @@ "issue:41720", "issue:42175", "issue:42915", + "issue:43257", "issue:43404", "issue:43452", "issue:43540", @@ -7121,6 +7132,7 @@ "issue:43582", "issue:43742", "issue:43749", + "issue:43756", "issue:44062", "issue:44291", "issue:44479", @@ -7131,7 +7143,6 @@ "issue:44792", "issue:44821", "issue:44849", - "issue:44857", "issue:44898", "issue:44936", "issue:44964", @@ -7148,33 +7159,31 @@ "issue:46512", "issue:46693", "issue:46821", - "issue:46832", "issue:46869", "issue:47020" ], "soft_pairs": [ - "issue:44857|issue:46832", - "issue:45081|issue:47020", "issue:43404|issue:45127", + "issue:45081|issue:47020", "issue:28282|issue:29127", - "issue:43742|issue:44291", "issue:44936|issue:46821", + "issue:43742|issue:44291", "issue:42915|issue:45538", "issue:44479|issue:46693", "issue:45538|issue:46869", "issue:44062|issue:44991", - "issue:43452|issue:43582", "issue:24643|issue:29159", "issue:27957|issue:28282", + "issue:43452|issue:43582", "issue:42915|issue:43749", "issue:43577|issue:46512", - "issue:41720|issue:43540", "issue:44849|issue:45701", + "issue:41720|issue:43540", "issue:42175|issue:44488", "issue:43540|issue:44560", "issue:44291|issue:44821", - "issue:43540|issue:44479", "issue:44987|issue:45907", + "issue:43540|issue:44479", "issue:34567|issue:36331", "issue:44479|issue:45381", "issue:44792|issue:44964", @@ -7182,222 +7191,230 @@ "issue:44062|issue:45081", "issue:44568|issue:45488", "issue:45042|issue:45081", - "issue:43742|issue:44589" + "issue:43257|issue:43756", + "issue:43742|issue:44589", + "issue:43582|issue:46821" ], - "prepared_review_unit_hash": "36e5bd84f47eb5870aaac87114f67e55d283c015642b331813164d734484ac48", + "prepared_review_unit_hash": "062a482a16f8e9db64c225ad8389612fb342cb2c92dd13bb953e8d117937c344", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36702, - "estimated_input_tokens": 9176, - "estimated_eval_tokens": 18608 + "soft_pair_count": 31, + "serialized_chars": 37011, + "estimated_input_tokens": 9253, + "estimated_eval_tokens": 18762 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is highly heterogeneous: most items are unrelated bugs across tokenization, training, multimodal loading, and device-mapping. Only one pair (the Qwen2.5-VL video regression reports) looks like the same underlying bug.", - "confidence": 0.89, - "canonical_issue_reason": "No safe canonical issue exists for the whole set; the reports span multiple unrelated code paths. If forced, issue 44479 is the most substantial standalone regression report, but it is not an umbrella for the rest.", + "summary": "This cluster is mostly heterogeneous: tokenizer regressions, model-loading failures, training bugs, and multimodal/video issues. Most soft pairs are not true duplicates. The only near-duplicate-looking subgroup is the Qwen video-input regression pair.", + "confidence": 0.56, + "canonical_issue_reason": "issue:44479 is the clearest umbrella regression report among the items, covering multiple Qwen video-input failures in v5.3.0 and plausibly subsuming the more specific Qwen2.5-VL symptom in 45381.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44479 is the best standalone representative: it describes a concrete versioned regression, names multiple affected models, and is specific enough to be actionable.", + "best_issue_reason": "issue:45488 is the strongest overall issue anchor for the cluster: it is detailed, actively discussed, has inbound references, and describes a broad tokenizer regression with real downstream impact.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44857", - "right": "issue:46832", + "left": "issue:43404", + "right": "issue:45127", "accept": false, - "reason": "Different failures in different models: LW-DETR AMP crash vs RT-DETR feature-level mismatch." + "reason": "Both involve tied/loaded weights, but they are different bugs: Mistral3 lm_head tying vs LoRA merge with extended vocabulary." }, { "left": "issue:45081", "right": "issue:47020", "accept": false, - "reason": "Different tokenizer bugs: Mistral regex patch crash vs XLMRoberta Unigram vocab type error." - }, - { - "left": "issue:43404", - "right": "issue:45127", - "accept": false, - "reason": "Both involve weights/tie embeddings, but they are distinct bugs in different generation/LoRA flows." + "reason": "Different tokenizer code paths and failures: Mistral regex patch crash vs XLMRoberta Unigram init type error." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "Unrelated errors: missing PyTorch import vs LayoutLMv3 box-validation message clarity." + "reason": "Unrelated problems: missing PyTorch import vs a LayoutLMv3 error-message clarity issue." }, { - "left": "issue:43742", - "right": "issue:44291", + "left": "issue:44936", + "right": "issue:46821", "accept": false, - "reason": "Different loading problems with different causes: key error on MobileLLM vs unexpected init_empty_weights argument." + "reason": "Trainer lifecycle bug vs TimesFM window_size AttributeError; no shared concrete code path." }, { - "left": "issue:44936", - "right": "issue:46821", + "left": "issue:43742", + "right": "issue:44291", "accept": false, - "reason": "Trainer lifecycle bug vs TimesFM window_size attribute error; no shared code path." + "reason": "Both are loading-time errors, but the causes differ: MobileLLM key error vs init_empty_weights unexpected argument." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "Different subsystems: Qwen3Moe FP8 config failure vs CLIP tokenizer max_length constant." + "reason": "FineGrainedFP8/Qwen3Moe failure is unrelated to CLIPTokenizer model_max_length behavior." }, { "left": "issue:44479", "right": "issue:46693", "accept": false, - "reason": "Both mention generation regressions, but one is a video-input bug and the other is a FA2 performance regression." + "reason": "Video-input regression vs FA2 long-context generation performance regression; different bug class." }, { "left": "issue:45538", "right": "issue:46869", "accept": false, - "reason": "Completely unrelated: tokenizer max_length vs YOLO multi-GPU device-map failure." + "reason": "Tokenizer max-length constant issue vs multi-GPU YOLO device-map failure." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "AddedToken special-arg duplication is a different tokenizer bug from EMBEDDIA/est-roberta loading." - }, - { - "left": "issue:43452", - "right": "issue:43582", - "accept": false, - "reason": "Different domains: GGUF/model loading vs Apple Silicon allocator warmup TypeError." + "reason": "Different tokenizer failures: AddedToken keyword collision vs loading EMBEDDIA/est-roberta." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed training weight-shape error vs slow/fast tokenizer inconsistency are unrelated." + "reason": "DeepSpeed training weight-shape runtime error is unrelated to slow/fast tokenizer inconsistency." }, { "left": "issue:27957", "right": "issue:28282", "accept": false, - "reason": "Flash Attention 2 on XLMRoberta is unrelated to PyTorch-not-found import errors." + "reason": "Flash Attention 2 on XLMRoberta is unrelated to AutoModel missing-PyTorch import handling." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "GGUF loading breakage and AppleSilicon caching allocator TypeError are different subsystems." }, { "left": "issue:42915", "right": "issue:43749", "accept": false, - "reason": "Both are Qwen3Moe/FSDP-related, but they describe different failures and code paths." + "reason": "Qwen3Moe FP8 config failure does not match the FSDP CPU RAM efficient loading regression." }, { "left": "issue:43577", "right": "issue:46512", "accept": false, - "reason": "Both concern dtype handling, but one is Blip2 loading dtypes and the other is stale config.dtype during from_config." + "reason": "Both concern dtype propagation, but one is Blip2 loading state and the other is from_config stale config metadata; not the same concrete bug." }, { - "left": "issue:41720", - "right": "issue:43540", + "left": "issue:44849", + "right": "issue:45701", "accept": false, - "reason": "Qwen3 auto device-mapping crash is distinct from Qwen3OmniMoe video-processing ValueError." + "reason": "Qwen3.5 hidden-states bug is unrelated to a broad tokenization-change report." }, { - "left": "issue:44849", - "right": "issue:45701", + "left": "issue:41720", + "right": "issue:43540", "accept": false, - "reason": "Qwen3.5 output_hidden_states bug is unrelated to a tokenizer behavior change across versions." + "reason": "Different Qwen issues: auto device mapping cuda assert vs video-input processing failure." }, { "left": "issue:42175", "right": "issue:44488", "accept": false, - "reason": "Packaging/backend dependency issue vs model-loading failure for cjvt/sleng-bert." + "reason": "Packaging/backend dependency issue is unrelated to a model-loading failure for cjvt/sleng-bert." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Both are video-input issues, but they affect different Qwen variants with different observed failures." + "reason": "Both are video-related, but they affect different models and the evidence does not show the same root bug." }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "init_empty_weights argument handling is unrelated to loading AutoImageProcessor from URL." + "reason": "Unexpected init_empty_weights argument and loading AutoImageProcessor from URL are distinct failures." }, { - "left": "issue:43540", - "right": "issue:44479", + "left": "issue:44987", + "right": "issue:45907", "accept": false, - "reason": "Shared video theme only; the reported failures and affected models differ." + "reason": "Model loading failure for a repo is unrelated to list_repo_files not forwarding revision." }, { - "left": "issue:44987", - "right": "issue:45907", + "left": "issue:43540", + "right": "issue:44479", "accept": false, - "reason": "Model loading failure vs list_repo_files revision-forwarding CI break are unrelated." + "reason": "Same general video area, but the model family and failure mode are different enough that the shared root bug is not established." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "TrainerState token accounting bug vs custom compute_loss signature break are different trainer issues." + "reason": "TrainerState token counter bug is unrelated to CustomTrainer.compute_loss signature changes." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "These look like the same Qwen2.5-VL video regression: both describe incorrect video position handling / vision_position_ids in the same family and version range." + "reason": "Both describe the same Qwen video-input regression in the 5.3.0 timeframe; 45381 looks like a specific symptom of the broader 44479 bug." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Different multimodal models and symptoms: janus image-generation test failure vs Phi-4-multimodal load failure." + "reason": "Janus image-generation test failure and Phi-4 multimodal loading failure are different issues." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 is unrelated to Perceiver non-default resolution failures." + "reason": "Unexpected zero probabilities in SigLIP2 is not the same as Perceiver image-classification resolution handling." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Different tokenizer bugs: AddedToken keyword collision vs Mistral regex patch backend-tokenizer access." + "reason": "AddedToken special-argument collision and Mistral regex patch crash are separate tokenizer bugs." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Both involve tokenizer behavior in v5, but one is add_special_tokens/BOS-EOS and the other is ByteLevel vs Metaspace pre-tokenizer override." + "reason": "Both are tokenizer regressions, but one is add_special_tokens BOS/EOS behavior and the other is LlamaTokenizer pre-tokenizer override; different code paths." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "PIL image-processor dependency regression is unrelated to Mistral tokenizer regex patching." + "reason": "PIL backend requiring torchvision is unrelated to the Mistral tokenizer regex crash." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MOE fine-grained FP8 loading and Smollm3 RoPE-layer dropping are unrelated model-specific bugs." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "Key error while loading MobileLLM and missing Float8 storage are different loading/runtime issues." + "reason": "MobileLLM key error and missing Float8 storage type are different loading/runtime failures." + }, + { + "left": "issue:43582", + "right": "issue:46821", + "accept": false, + "reason": "AppleSilicon allocator warmup TypeError is unrelated to the TimesFM window_size AttributeError." } ] }, "evaluator_result": { "accept": true, - "feedback": "The packet supports the overall conclusion: the set is heterogeneous, and the only clearly mergeable soft pair is issue:44479 with issue:45381, which both describe the same Qwen2.5-VL video-input regression around vision_position_ids. The other false pair judgments are conservative and grounded in distinct titles/symptoms. No obvious overstatement or unsupported canonical claim stands out." + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The only accepted pair, 44479/45381, is plausibly the same Qwen video-input regression family, while the rest are reasonably rejected as distinct bugs. The chosen anchors are also supported by the activity/inbound-reference signals in the packet." }, "evaluator_used": true, "retried": false, @@ -7423,15 +7440,14 @@ "issue:29870", "issue:32090", "issue:41762", - "issue:43257", "issue:43452", "issue:43526", "issue:43531", "issue:43540", "issue:43577", - "issue:43582", "issue:43697", "issue:43716", + "issue:43720", "issue:43756", "issue:43906", "issue:44488", @@ -7440,6 +7456,7 @@ "issue:44737", "issue:44821", "issue:44877", + "issue:44898", "issue:44913", "issue:44936", "issue:45072", @@ -7462,26 +7479,23 @@ "issue:46491", "issue:46693", "issue:46701", - "issue:46821", "issue:46829", "issue:46846", "issue:46899" ], "soft_pairs": [ - "issue:43257|issue:43756", - "issue:43582|issue:46821", "issue:43531|issue:44821", "issue:43756|issue:46399", "issue:43756|issue:45325", - "issue:44737|issue:46463", "issue:45405|issue:45600", + "issue:44737|issue:46463", "issue:43716|issue:46701", - "issue:43577|issue:45072", - "issue:45823|issue:46463", "issue:44568|issue:46491", + "issue:45823|issue:46463", "issue:24643|issue:29870", - "issue:45672|issue:45674", + "issue:43577|issue:45072", "issue:43906|issue:46693", + "issue:45672|issue:45674", "issue:44625|issue:44877", "issue:43452|issue:44488", "issue:43540|issue:46693", @@ -7489,210 +7503,205 @@ "issue:43906|issue:46468", "issue:43697|issue:46829", "issue:44913|issue:46121", - "issue:44737|issue:45588", "issue:45663|issue:46846", + "issue:44737|issue:45588", "issue:43526|issue:45479", "issue:46463|issue:46899", "issue:43697|issue:45216", + "issue:41762|issue:44936", "issue:45357|issue:45920", - "issue:41762|issue:44936" + "issue:43720|issue:44898" ], - "prepared_review_unit_hash": "c5337b40f2dc1246a2f0ad9e9c6263ac2ee3163a9020eb91fc90c83af45fbb09", + "prepared_review_unit_hash": "236fc104afbc0e918412f6cd5798e401f1e22aec75e9d1c973017af6aa527658", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36669, - "estimated_input_tokens": 9168, - "estimated_eval_tokens": 18592 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35864, + "estimated_input_tokens": 8966, + "estimated_eval_tokens": 18188 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft pairs are all near-matches by wording/theme only, not the same underlying bug or change. There is no convincing duplicate canonical issue/PR across the set.", - "confidence": 0.71, - "canonical_issue_reason": null, + "summary": "These soft matches are mostly keyword-level overlaps (dtype, load, config, generation) across unrelated models and code paths. None of the candidate pairs look like the same underlying bug/change, so there is no clear canonical issue in this cluster.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue: the items are heterogeneous and the soft-similarity pairs do not converge on one shared defect.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46693 is the strongest representative of the cluster only because it has the highest activity and inbound references, but it is not a true duplicate canonical for the whole set.", + "best_issue_reason": "No strong global best issue: none of these issues is representative enough to serve as a reliable cluster anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43257", - "right": "issue:43756", - "accept": false, - "reason": "Different models and failure modes: Qwen3 MoE weight conversion under accelerate/deepspeed vs Smollm3 RoPE layer dropping." - }, - { - "left": "issue:43582", - "right": "issue:46821", - "accept": false, - "reason": "Unrelated runtime paths: Apple Silicon allocator warmup TypeError vs TimesFM window_size AttributeError." - }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor-from-URL loading are different bugs." + "reason": "Different features and failures: Qwen3-MoE sliding-window behavior vs loading AutoImageProcessor from a URL." }, { "left": "issue:43756", "right": "issue:46399", "accept": false, - "reason": "Both mention RoPE-ish model internals, but one is Smollm3 layer handling and the other is a Qwen3.5 kernel registration issue." + "reason": "Both mention model internals, but the bugs are unrelated: Smollm3 RoPE-layer dropping vs Qwen3.5 kernel registration failure." }, { "left": "issue:43756", "right": "issue:45325", "accept": false, - "reason": "Same broad RoPE family, but distinct model families and distinct errors; not one concrete bug." + "reason": "Different models and different bugs: RoPE-layer selection in Smollm3 vs Qwen2.5-VL temporal position_id scaling." }, { - "left": "issue:44737", - "right": "issue:46463", + "left": "issue:45405", + "right": "issue:45600", "accept": false, - "reason": "XLNet CPU-side relative position encoding bug is unrelated to T5 decoder input validation messaging." + "reason": "A dependency-version bump is unrelated to removing stale auto-mapping entries; no shared code-path bug." }, { - "left": "issue:45405", - "right": "issue:45600", + "left": "issue:44737", + "right": "issue:46463", "accept": false, - "reason": "A dependency version bump and a stale auto_mappings reference are different maintenance issues." + "reason": "XLNet CPU tensor creation bug vs T5 decoder-input validation error message; unrelated subsystems." }, { "left": "issue:43716", "right": "issue:46701", "accept": false, - "reason": "Both are multimodal failures, but the dtype mismatch in Mistral-3 is not the same as the DiffusionGemma pipeline breakage." + "reason": "Different multimodal pipelines and failure modes: dtype mismatch in Mistral-3 image preprocessing vs broken DiffusionGemma image-text-to-text generation." }, { - "left": "issue:43577", - "right": "issue:45072", + "left": "issue:44568", + "right": "issue:46491", "accept": false, - "reason": "Both involve dtype mismatches, but in different models and code paths; they would not be fixed by one change." + "reason": "Tokenizer behavior differs: one is missing BOS/EOS insertion, the other is a leading-space round-trip regression in CodeLlama." }, { "left": "issue:45823", "right": "issue:46463", "accept": false, - "reason": "Gemma4 device-map forward mismatch is unrelated to T5 decoder input validation." - }, - { - "left": "issue:44568", - "right": "issue:46491", - "accept": false, - "reason": "Tokenizer regressions, but one is BOS/EOS insertion for mdeberta and the other is leading-space round-trip behavior for CodeLlama." + "reason": "Gemma4 device-placement mismatch during forward is unrelated to T5 decoder-input validation." }, { "left": "issue:24643", "right": "issue:29870", "accept": false, - "reason": "DeepSpeed training weight-dimension error and Bart generate() UnboundLocalError are unrelated failures." + "reason": "DeepSpeed training weight-shape error and Bart generate() unbound local variable are different bugs in different execution paths." }, { - "left": "issue:45672", - "right": "issue:45674", + "left": "issue:43577", + "right": "issue:45072", "accept": false, - "reason": "Both touch 4-bit inference, but one is unsafe dtype casting in Gemma4 audio/vision embedders and the other is a BitsAndBytesConfig skip-module exclusion bug." + "reason": "Both are dtype-related, but they affect different models and causes: Blip2 loading dtypes vs SwitchTransformers/TimmWrapper bfloat16 inference mismatches." }, { "left": "issue:43906", "right": "issue:46693", "accept": false, - "reason": "An isolated reproduction of an older issue is not the same as the FA2 long-context performance regression." + "reason": "An isolated reproduction of an older issue is not the same as a FA2 long-context performance regression." + }, + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Both involve quantization, but one is an unsafe dtype cast in Gemma4 audio/vision embedders and the other is a BitsAndBytes skip-module exclusion-list bug." }, { "left": "issue:44625", "right": "issue:44877", "accept": false, - "reason": "Qwen3.5 num_labels propagation and strict granite_speech config loading are different config-loading problems." + "reason": "Qwen3.5 num_labels propagation and strict granite_speech config loading are different configuration bugs." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "Both are model-loading complaints, but the gguf_file/from_pretrained breakage is not the same bug as failing to load cjvt/sleng-bert." + "reason": "Both are loading failures, but the affected objects and root causes differ: gguf_file handling vs cjvt/sleng-bert model loading." }, { "left": "issue:43540", "right": "issue:46693", "accept": false, - "reason": "Video-input processing error in Qwen3OmniMoe is unrelated to FA2 long-context generation slowdown/regression." + "reason": "Video input processing in Qwen3OmniMoe is unrelated to FA2 long-context generation performance regression." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Missing PyTorch import and _gpu_broadcast_one(NoneType) are unrelated trainer/environment issues." + "reason": "PyTorch import failure and _gpu_broadcast_one NoneType error are unrelated environment/runtime failures." }, { "left": "issue:43906", "right": "issue:46468", "accept": false, - "reason": "A reproduction of issue 38071 is not the same as the broken Nemotron example." + "reason": "The Nemotron example breakage is not the same as the isolated reproduction of issue 38071." }, { "left": "issue:43697", "right": "issue:46829", "accept": false, - "reason": "Different model outputs and different bug classes: RTDetrV2 regression vs VideoPrism output field mix-up." + "reason": "RTDetrV2 output differences and VideoPrism hidden_states misassignment are unrelated model-specific bugs." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "Both involve rotary/rope configuration, but one is silent config reset on reload and the other is JSON-list type handling." + "reason": "Both involve rope-related config handling, but one is rotary_pct not persisting on reload and the other is a TypeError from JSON-loaded ignore_keys_at_rope_validation." }, { - "left": "issue:44737", - "right": "issue:45588", + "left": "issue:45663", + "right": "issue:46846", "accept": false, - "reason": "XLNet position encoding on CPU is unrelated to flash_attention's s_aux=None crash." + "reason": "Gemma4 FSDP2 shared_kv_states rebuilding and Qwen3.5 tensor-parallel plan omissions are different distributed-training bugs." }, { - "left": "issue:45663", - "right": "issue:46846", + "left": "issue:44737", + "right": "issue:45588", "accept": false, - "reason": "Gemma4 FSDP2 shared_kv_states KeyError and Qwen3.5 TP-plan omissions are separate distributed-training bugs." + "reason": "XLNet CPU placement bug and flash_attention s_aux=None crash are unrelated code paths." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "BeitImageProcessorFast label reduction and zero-loss single-label classification are different label-handling problems." + "reason": "BeitImageProcessorFast label reduction and degenerate zero-loss sequence classification are different issues in different components." }, { "left": "issue:46463", "right": "issue:46899", "accept": false, - "reason": "T5 decoder validation messaging and Gemma4 4-bit multimodal dtype casting are unrelated." + "reason": "T5 decoder-input validation messaging is unrelated to Gemma4 4-bit audio/vision dtype casting." }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "RTDetrV2 inference regression and Qwen3.5 save_pretrained checkpoint corruption are unrelated." + "reason": "Different regressions: RTDetrV2 output drift vs Qwen3.5 incorrect save_pretrained checkpoint contents." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading index error is unrelated to trainer.evaluate() failing after train()." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Qwen3.5 visual-encoder key saving and tokenizer token-id regressions are different serialization/tokenization bugs." + "reason": "Qwen3.5 visual-encoder checkpoint saving and tokenizer token-ID regressions are different bugs." }, { - "left": "issue:41762", - "right": "issue:44936", + "left": "issue:43720", + "right": "issue:44898", "accept": false, - "reason": "Gemma3 ZeRO-3 load IndexError and trainer.evaluate-after-train failure are not the same code path." + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific problems." } ] }, @@ -7727,7 +7736,6 @@ "issue:43646", "issue:43653", "issue:43688", - "issue:43720", "issue:43742", "issue:43749", "issue:43756", @@ -7767,9 +7775,8 @@ ], "soft_pairs": [ "issue:44466|issue:45005", - "issue:43720|issue:44898", - "issue:43643|issue:45375", "issue:45200|issue:46463", + "issue:43643|issue:45375", "issue:44493|issue:46050", "issue:44361|issue:44442", "issue:43688|issue:43749", @@ -7778,229 +7785,220 @@ "issue:43452|issue:44843", "issue:43531|issue:43742", "issue:43819|issue:44811", - "issue:43761|issue:46829", "issue:43994|issue:44898", - "issue:45372|issue:46170", + "issue:43761|issue:46829", "issue:44857|issue:47000", "issue:29159|issue:33357", - "issue:45542|issue:46178", + "issue:45372|issue:46170", "issue:43525|issue:46701", + "issue:45542|issue:46178", "issue:46832|issue:46932", - "issue:43295|issue:45356", "issue:43688|issue:45704", + "issue:43295|issue:45356", "issue:43688|issue:45701", "issue:43653|issue:44568", - "issue:43646|issue:43749", "issue:43756|issue:46402", + "issue:43646|issue:43749", "issue:44464|issue:44743", "issue:43653|issue:43927" ], - "prepared_review_unit_hash": "fb48e84abfafd6553f3994670d9d89361c94ec105994fdbdd405f6201945f982", + "prepared_review_unit_hash": "30c1f592a48e4e10447b4411f643577d1ac5a887a27c06f5e8081b95c8d05f44", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36385, - "estimated_input_tokens": 9097, - "estimated_eval_tokens": 18450 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35483, + "estimated_input_tokens": 8871, + "estimated_eval_tokens": 17998 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is still mostly heterogeneous. There is one strong duplicate pair around tied-weights serialization regressions, but the Windows AMD ROCm import issue is unrelated to the Gemma 4 / mistral_common problem, so the cluster should not be treated as a single duplicate set.", - "confidence": 0.88, - "canonical_issue_reason": "Issue 29159 is the broadest representative issue in the packet because it describes a core tokenizer behavior inconsistency, but it does not unify the cluster into one underlying bug.", + "summary": "The cluster is highly heterogeneous, and none of the soft pairs look like the same underlying bug with enough confidence to merge. I would keep the issues separate; there is also no PR content here to evaluate.", + "confidence": 0.84, + "canonical_issue_reason": "No single canonical issue cleanly represents the cluster: most items are unrelated regressions or model-specific bugs. If an anchor is needed, 44466 is the closest because it is the most detailed tied-weights report.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 29159 is the best single anchor for the cluster because it is the most general and central issue among otherwise unrelated bugs.", + "best_issue_reason": "44466 is the strongest standalone report in the set, but it only covers a tied-weight serialization regression and does not subsume the rest of the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44466", "right": "issue:45005", - "accept": true, - "reason": "Both describe the same v5 tied-weights serialization regression affecting translation models and lm_head weight handling." + "accept": false, + "reason": "Both mention tied weights in v5, but one is a device-dependent serialization issue and the other is a broader translation-model complaint; not clearly the same bug." }, { - "left": "issue:43720", - "right": "issue:44898", + "left": "issue:45200", + "right": "issue:46463", "accept": false, - "reason": "Different bugs: BitNet accelerate loading versus Perceiver image-classification resolution/interpolation failure." + "reason": "Different failures: Gemma 4 mm_token_type_ids behavior vs T5 decoder-input validation messaging." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both concern missing config fields, but they target different model families and different code paths." - }, - { - "left": "issue:45200", - "right": "issue:46463", - "accept": false, - "reason": "Unrelated issues: Gemma 4 mm_token_type_ids defaults versus T5 decoder-input validation messaging." + "reason": "Both involve config fields being missing, but the causes differ: trust_remote_code AutoConfig behavior vs @strict dropping a specific Qwen field." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Different regressions: unexpected position-id keys versus quantized model save/load breakage." + "reason": "Different symptoms and paths: unexpected position-id keys vs quantized model save/load breakage." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Both are loading failures, but for different tokenizer classes and different underlying causes." + "reason": "Both are tokenizer load errors, but they affect different tokenizers and likely different code paths." }, { "left": "issue:43688", "right": "issue:43749", "accept": false, - "reason": "MoE auxiliary-loss normalization is unrelated to FSDP CPU RAM efficient loading." + "reason": "Aux-loss normalization in MoE models is unrelated to an FSDP CPU RAM-efficient loading regression." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "Different tokenizer bugs with different symptoms and code paths." + "reason": "BigBird special-token registration is unrelated to MLukeTokenizer task AttributeErrors." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "SigLIP zero-probability outputs are not the same as the processor.tokenizer regression." + "reason": "A zero-probability SigLIP model output is not the same bug as a processor/tokenizer regression in custom model code." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "gguf_file loading breakage and offline model_info calls are separate loader issues." + "reason": "Both are loading-related, but gguf_file breakage and an unconditional model_info() call in offline mode are distinct issues." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior and MobileLLM key errors are distinct problems." + "reason": "Different model families and failures: Qwen3-MoE sliding_window handling vs a key error loading MobileLLM-125M." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC latent reconstruction and Whisper batch_decode special-token handling are different subsystems." - }, - { - "left": "issue:43761", - "right": "issue:46829", - "accept": false, - "reason": "Different model families and different output-field plumbing problems." + "reason": "DAC latent reconstruction mismatch and Whisper batch_decode skip_special_tokens handling are unrelated code paths." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Distinct vision-model failures with different symptoms and likely causes." + "reason": "Both are vision-model regressions, but one is nonsensical SigLIP outputs and the other is Perceiver resolution/interpolation failure." }, { - "left": "issue:45372", - "right": "issue:46170", + "left": "issue:43761", + "right": "issue:46829", "accept": false, - "reason": "Not the same bug: Gemma 4 processor loading fails on a mistral_common ReasoningEffort import, while this is a Windows AMD ROCm / torch.distributed dependency issue." + "reason": "Both mention hidden_states, but the models and the failure modes differ: None returned vs wrong tensor placed in the output." }, { "left": "issue:44857", "right": "issue:47000", "accept": false, - "reason": "Different DETR loss crashes with different failure conditions." + "reason": "Different losses and different crashes: float16 AMP failure in LwDetrImageLoss vs infeasible cost matrix in RTDetrLoss." }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to a MacOS CLIP bus error." + "reason": "Tokenizer inconsistency is not the same as a MacOS bus error in CLIP model loading." }, { - "left": "issue:45542", - "right": "issue:46178", + "left": "issue:45372", + "right": "issue:46170", "accept": false, - "reason": "Both are backend/dependency errors, but they involve different missing backends and different mechanisms." + "reason": "Separate import/dependency problems: Gemma 4 processor loading vs Windows ROCm torch.distributed dependency issues." }, { "left": "issue:43525", "right": "issue:46701", "accept": false, - "reason": "Llama4Config pad_token_id and DiffusionGemma pipeline failures are unrelated." + "reason": "Llama4Config missing pad_token_id and DiffusionGemma pipeline failure are unrelated." }, { - "left": "issue:46832", - "right": "issue:46932", + "left": "issue:45542", + "right": "issue:46178", "accept": false, - "reason": "RT-DETR feature-level shape error and Gemma4 quantized-embedder dtype casting are separate issues." + "reason": "Both are backend/import issues, but they involve different backends and different failure mechanisms." }, { - "left": "issue:43295", - "right": "issue:45356", + "left": "issue:46832", + "right": "issue:46932", "accept": false, - "reason": "Different regressions: hidden_states handling versus tokenizer codec / regex behavior." + "reason": "RTDetrModel feature-level mismatch and Gemma4 quantized embedder dtype casting are unrelated bugs." }, { "left": "issue:43688", "right": "issue:45704", "accept": false, - "reason": "MoE auxiliary-loss normalization is unrelated to a T5 apex RMSNorm memory leak." + "reason": "Aux-loss normalization in MoE models is unrelated to a T5/apex memory leak." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor/tokenizer image-handling regression and Kimi-K2.5 codec/mistral-regex regression are different issues." }, { "left": "issue:43688", "right": "issue:45701", "accept": false, - "reason": "Aux-loss normalization and version-dependent tokenization are different bugs." + "reason": "Aux-loss normalization has nothing to do with a report that transformers version changes tokenization." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Different tokenizer special-token issues affecting different tokenizers." + "reason": "BigBird mask-token registration and mdeberta-v3 BOS/EOS insertion are separate tokenizer-special-token bugs." }, { - "left": "issue:43646", - "right": "issue:43749", + "left": "issue:43756", + "right": "issue:46402", "accept": false, - "reason": "Custom model initialization breakage is not the same as FSDP CPU RAM efficient loading." + "reason": "Smollm3 RoPE-layer dropping and Sam3Model text-encoder weight loading are unrelated model-specific regressions." }, { - "left": "issue:43756", - "right": "issue:46402", + "left": "issue:43646", + "right": "issue:43749", "accept": false, - "reason": "RoPE layer selection in Smollm3 is unrelated to Sam3Model text encoder loading." + "reason": "Custom model initialization breaks under v5, while the other report is about FSDP CPU RAM-efficient loading." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Chunked generation with compiled forward and Qwen3_5 recurrent-state resets are different code-path bugs." + "reason": "Chunked generation with compiled forward and Qwen3_5MoE recurrent-state resetting are different code paths." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "BigBird mask-token registration and DiaConfig token-ID persistence are unrelated." + "reason": "BigBird special-token registration and DiaConfig custom token ID persistence are unrelated tokenizer/config persistence bugs." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The summary is not fully grounded in the packet. In particular, the accepted soft-edge pair issue:45372 vs issue:46170 is not conservative: 45372 is about a mistral_common `ReasoningEffort` import blocking Gemma 4 processor loading, while 46170 is a Windows AMD ROCm / `torch.distributed` dependency issue. Those look unrelated, so the claim of a clear duplicate pair overstates the evidence. The rest is mostly conservative, but this one false acceptance means the cluster verdict should be rejected." - }, - "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -8069,9 +8067,9 @@ "soft_pairs": [ "issue:44821|issue:46821", "issue:42898|issue:46491", - "issue:43452|issue:43577", "issue:44843|issue:44991", "issue:44479|issue:44625", + "issue:43452|issue:43577", "issue:33666|issue:36010", "issue:43761|issue:47013", "issue:32090|issue:33666", @@ -8086,226 +8084,226 @@ "issue:33666|issue:36331", "issue:43278|issue:43688", "issue:43504|issue:43720", - "issue:45491|issue:46002", "issue:43782|issue:46693", "issue:43720|issue:43927", "issue:43577|issue:44554", - "issue:43611|issue:46846", "issue:43606|issue:43927", + "issue:43611|issue:46846", "issue:43525|issue:43572", "issue:43504|issue:44610", "issue:43720|issue:44361", + "issue:45491|issue:46002", "issue:43643|issue:44561", "issue:44964|issue:45405" ], - "prepared_review_unit_hash": "8872d70920aa6758e67c3894586f19357f74a25feb735fccc9e2964003e15acc", + "prepared_review_unit_hash": "16fda82bb35510e7217b43b4dd84c6936326b0c17afb3d71f0e4cc622953242d", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36843, - "estimated_input_tokens": 9211, - "estimated_eval_tokens": 18678 + "serialized_chars": 36847, + "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 highly heterogeneous: it mixes unrelated tokenizer, config, loading, generation, training, and device-mapping bugs. None of the soft pairs look like the same underlying defect, so I would not merge any of them as duplicates.", - "confidence": 0.92, - "canonical_issue_reason": "No clear canonical issue: the items are a grab-bag of distinct regressions and model-specific bugs rather than one underlying problem.", + "summary": "The cluster is heterogeneous: it mixes unrelated tokenizer, config, loading, training, generation, and performance regressions across different models and code paths. None of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue stands out: the items span unrelated model-specific regressions and API breakages rather than one shared defect.", "canonical_pr_reason": null, - "best_issue_reason": "No single best issue for dedup triage; the soft-linked items do not form a true duplicate set.", + "best_issue_reason": "No global best issue: there isn\u2019t a representative issue that subsumes the others, and the apparent similarity pairs still break on different code paths and causes.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44821", "right": "issue:46821", "accept": false, - "reason": "Different subsystems: image processor URL loading vs TimesFM window_size attribute handling." + "reason": "Different failures: URL loading for `AutoImageProcessor` vs a TimesFM `window_size` attribute error." }, { "left": "issue:42898", "right": "issue:46491", "accept": false, - "reason": "Both concern tokenization behavior, but one is v5 default cleanup semantics and the other is CodeLlama leading-space round-trip regression." - }, - { - "left": "issue:43452", - "right": "issue:43577", - "accept": false, - "reason": "Unrelated bugs: gguf_file/from_pretrained loading vs BF16/FP32 dtype mismatch in Blip2." + "reason": "Both tokenizer-related, but one is a v5 `clean_up_tokenization_spaces` behavior change and the other is a CodeLlama round-trip spacing regression." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Both involve tokenizer loading, but one is an offline hub regex patch issue and the other is a specific tokenizer failure on a model repo." + "reason": "One is an offline-mode break from `_patch_mistral_regex` calling `model_info()`, the other is a tokenizer load failure for a different model." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Same Qwen family but different problems: video-input regression vs num_labels propagation in config." + "reason": "Qwen video-input regression vs missing `num_labels` propagation; different models and different bugs." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "`gguf_file` loading breakage vs BLIP2 dtype drift during load; unrelated code paths." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training support issue vs an import break for GenerationMixin; different failure modes and code paths." + "reason": "Multi-GPU training issue vs `GenerationMixin` import error; not the same defect." }, { "left": "issue:43761", "right": "issue:47013", "accept": false, - "reason": "Different model/runtime bugs: CLIP vision hidden_states handling vs xLSTM return shape mismatch." + "reason": "CLIP hidden-states regression vs xLSTM train-mode tuple unpacking crash; different models and mechanics." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Both mention Trainer/multi-GPU, but one is None broadcast error and the other is a general Qwen2-VL multi-GPU training request/problem." + "reason": "Both involve training, but one is `_gpu_broadcast_one` on `NoneType` and the other is a Qwen2-VL multi-GPU training report." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "Different serialization bugs: device-dependent lm_head.weight saving vs sinq quantized model save/load breakage." + "reason": "Different serialization bugs: tied-weight serialization varies by device vs quantized `sinq` save/load breakage." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "Unrelated paths: save_pretrained shard-size question vs AutoProcessor kwargs not forwarded to cached_file." + "reason": "`max_shard_size` default question vs `AutoProcessor.from_pretrained` not forwarding kwargs to `cached_file`." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Both are MoE-related, but one is auxiliary-loss normalization and the other is DeepseekV3 implementation divergence." + "reason": "Both MoE-related, but one is auxiliary-loss normalization and the other is native vs remote DeepSeekV3 implementation divergence." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Different bugs: auto device mapping assert on A800 vs compiled-forward chunked generation inconsistency." + "reason": "CUDA assert with auto device mapping vs chunked-generation inconsistency under compiled forward; different symptoms and paths." }, { "left": "issue:44977", "right": "issue:46693", "accept": false, - "reason": "Both mention flash-attention/long-context generation, but one is Qwen3.5 correctness and the other is a broader performance regression." + "reason": "Qwen3.5 flash-attention correctness issue vs FA2 long-context performance regression; not the same bug." }, { "left": "issue:44464", "right": "issue:44977", "accept": false, - "reason": "Compiled-forward chunked generation bug vs flash-attention generation issue; not the same code-path." + "reason": "Compiled-forward chunked generation bug vs flash-attention generation failure; different triggers and code paths." }, { "left": "issue:44734", "right": "issue:46612", "accept": false, - "reason": "Both involve generation cache behavior, but one is KV-cache continuation indexing and the other is beam-search reorder skipping for certain model classes." + "reason": "KV-cache continuation indexing bug in serve vs beam-search cache reorder being skipped for several recurrent models." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Different model loading vs compile/runtime issues: Beit legacy-field load failure vs Bamba SDPA torch.compile failure." + "reason": "Legacy-field model loading failure vs Bamba SDPA/torch.compile failure; unrelated model behaviors." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Trainer-related in a broad sense, but one is multi-GPU training and the other is a compute_loss signature/API break." + "reason": "Both touch training, but one is multi-GPU setup and the other is a `compute_loss()` signature incompatibility." }, { "left": "issue:43278", "right": "issue:43688", "accept": false, - "reason": "Different defect types: dtype drift between train/eval vs MoE auxiliary-loss normalization." + "reason": "Embedding dtype drift in eval vs MoE auxiliary-loss normalization; no shared underlying bug." }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Both are loading-related, but one is a legacy field in Beit and the other is BitNet packed-weight unpacking under accelerate." - }, - { - "left": "issue:45491", - "right": "issue:46002", - "accept": false, - "reason": "Both produce NaNs, but the triggering conditions and components differ: Gemma3 sliding-window attention vs classification head init/checkpointing." + "reason": "Beit legacy-field loading issue vs BitNet packed-weight unpacking during accelerate loading." }, { "left": "issue:43782", "right": "issue:46693", "accept": false, - "reason": "Qwen3VL load error with weight_only vs general FA2 long-context performance regression; unrelated." + "reason": "Qwen3VL `weight_only=True` load error vs FA2 performance regression; different failure classes." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "Different loading bugs: packed weights not unpacked during accelerate loading vs DiaConfig losing custom token IDs on save/load." + "reason": "BitNet accelerate-loading bug vs DiaConfig token-ID loss after save/load; unrelated." }, { "left": "issue:43577", "right": "issue:44554", "accept": false, - "reason": "Unrelated: Blip2 dtype propagation vs MPS attention correctness when value head dim differs from query." + "reason": "BLIP2 dtype propagation vs MPS attention correctness when value/query head dims differ." }, { - "left": "issue:43611", - "right": "issue:46846", + "left": "issue:43606", + "right": "issue:43927", "accept": false, - "reason": "Different failures: base_model_prefix loading breakage in v5 vs Qwen3.5 tensor-parallel plan missing linear_attn entries." + "reason": "CPU-offload device mismatch vs config token-ID loss and generation `IndexError`; different code paths." }, { - "left": "issue:43606", - "right": "issue:43927", + "left": "issue:43611", + "right": "issue:46846", "accept": false, - "reason": "Device-mismatch CPU offload bug vs config token-ID loss on save/load; no shared underlying defect." + "reason": "Base-model-prefix loading regression vs Qwen3.5 TP-plan missing `linear_attn` entries." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Both are missing config fields after v5 changes, but they affect different models and different attributes (pad_token_id vs pad_token_idx)." + "reason": "Similar theme, but different model configs and different missing fields (`pad_token_id` vs `pad_token_idx`)." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Beit pretrained-load preset issue vs OmDet-Turbo processor/model size mismatch; unrelated." + "reason": "Legacy-field loading failure vs OmDet-Turbo processor/model input-size mismatch." }, { "left": "issue:43720", "right": "issue:44361", "accept": false, - "reason": "BitNet accelerate loading bug vs MLukeTokenizer AttributeError on tasks; different components and root causes." + "reason": "Packed-weight loading issue vs MLukeTokenizer task-time AttributeError." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Gemma3 mixed-length GPU NaNs vs classification-head NaNs after checkout/init; different conditions and subsystems." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both affect trust_remote_code, but one is missing fields from AutoConfig and the other is removal of is_torch_fx_available." + "reason": "`AutoConfig` missing fields under `trust_remote_code` vs removal of `is_torch_fx_available` breaking remote-code models." }, { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "Completely unrelated: Phi-4 multimodal loading regression vs a PEFT minimum-version packaging issue." + "reason": "Phi-4 multimodal loading failure vs an unreleased PEFT version pin; unrelated." } ] }, @@ -8336,7 +8334,6 @@ "issue:42796", "issue:43278", "issue:43295", - "issue:43334", "issue:43452", "issue:43493", "issue:43572", @@ -8363,10 +8360,10 @@ "issue:44849", "issue:44977", "issue:45290", + "issue:45357", "issue:45405", "issue:45520", "issue:45584", - "issue:45759", "issue:45812", "issue:45823", "issue:46050", @@ -8387,236 +8384,230 @@ "issue:44554|issue:46365", "issue:43452|issue:44451", "issue:45823|issue:46050", - "issue:42796|issue:44977", "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:43278|issue:43701", + "issue:43646|issue:44466", "issue:46209|issue:46710", "issue:44568|issue:46489", "issue:43927|issue:44361", "issue:33357|issue:33666", "issue:44849|issue:46209", - "issue:43575|issue:45405", "issue:44792|issue:45405", + "issue:43575|issue:45405", "issue:44568|issue:45812", "issue:43572|issue:44291", "issue:44488|issue:44843", - "issue:46821|issue:46832", - "issue:43334|issue:45759", + "issue:43824|issue:44589", "issue:44589|issue:45520", - "issue:43824|issue:44589" + "issue:46821|issue:46832", + "issue:44849|issue:45357" ], - "prepared_review_unit_hash": "432cd7462e9b7aa6c5b49ab1cde1f0a2682212057ba6c8afc635452003663eca", + "prepared_review_unit_hash": "698d73496e2cdb96d7c67edd37b9dee402dfb49a7c66b67225999c62bb3dcd53", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36524, - "estimated_input_tokens": 9131, - "estimated_eval_tokens": 18518 + "serialized_chars": 35882, + "estimated_input_tokens": 8971, + "estimated_eval_tokens": 18198 }, "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 buzzwords like tokenizer, model loading, or training regression, but they describe different models, code paths, and failure modes. None of the candidate edges looks like a safe duplicate merge.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue is justified here; the items are not one underlying bug cluster but a set of unrelated reports with only superficial text similarity.", - "canonical_pr_reason": "No pull requests are present in this cluster.", - "best_issue_reason": "No issue is a strong global representative, because the cluster spans unrelated tokenizer, multimodal, loading, config, and training problems rather than one shared defect.", - "best_pr_reason": "No pull requests are present in this cluster.", + "summary": "This cluster is heterogeneous: most items are unrelated model, tokenizer, config, or infrastructure bugs. None of the proposed soft pairs look like the same underlying issue, so all soft edges are rejected. If one issue must represent the cluster, the broad Granite tokenizer regression is the strongest candidate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45812 is the strongest representative: it describes a broad, concrete tokenizer regression with clear user impact and inbound references, whereas the other issues are mostly narrow and unrelated.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the best issue candidate because it is a high-signal, broadly scoped tokenizer regression with a concrete failure mode and visible downstream impact.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "Different models and failure modes: Qwen3.5 label propagation vs Whisper empty-transcription generation after token alignment." + "reason": "Different bugs: Qwen3.5 num_labels propagation vs Whisper empty transcription after token alignment." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Both involve apply_chat_template crashes, but the triggers differ: batched padding=False input vs assistant tool-call messages with no content." + "reason": "Same API area, but different crashes: batched apply_chat_template with padding=False vs assistant tool-call messages with no content." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "Unrelated bugs: SigLIP2 implementation discrepancy vs an incorrect pipeline translation-task error message." + "reason": "Both mention model/pipeline behavior, but one is a SigLIP2 implementation discrepancy and the other is a pipeline translation error message issue." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Different code paths and symptoms: Qwen3.5 flash-attention generation issue vs FP8 inference failure on B200." + "reason": "Qwen3.5 flash-attention generation bug vs FP8 inference failure on B200; different code paths and symptoms." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer spacing bug and a MacOS bus error for a CLIP model are not the same underlying problem." + "reason": "Tokenizer spacing issue and a MacOS CLIP bus error are unrelated failures." }, { "left": "issue:44554", "right": "issue:46365", "accept": false, - "reason": "MPS attention correctness is unrelated to importing Trainer failing because of datasets." + "reason": "MPS attention correctness regression vs Trainer import failing because of datasets; unrelated." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "gguf_file handling in from_pretrained is a different issue from a specific model load failure." + "reason": "Both are loading problems, but gguf_file handling and loading a specific model are distinct root causes." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Device-map mismatch in Gemma4 PLE is unrelated to sinq quantization save/load breakage." + "reason": "Gemma4 device mismatch during forward is unrelated to sinq quantized model save/load breakage." }, { - "left": "issue:42796", - "right": "issue:44977", + "left": "issue:43994", + "right": "issue:46192", "accept": false, - "reason": "Tokenizer BPE marker handling and Qwen3.5 flash-attention generation are separate problems." + "reason": "SigLIP2 output corruption and return_tensors being ignored are different bugs." }, { - "left": "issue:43994", - "right": "issue:46192", + "left": "issue:42796", + "right": "issue:44977", "accept": false, - "reason": "SigLIP2 wrong outputs are unrelated to return_tensors being ignored when text_kwargs is passed." + "reason": "Tokenizer BPE marker handling and Qwen3.5 flash-attention generation are unrelated." }, { "left": "issue:39692", "right": "issue:43295", "accept": false, - "reason": "Both mention multimodal processor/tokenizer usage, but one is a docs example with model/processor mismatch and quantization failure; the other is a regression in processor.tokenizer/image handling." + "reason": "Documentation/example mistakes are not the same as the processor.tokenizer regression in custom model code." }, { "left": "issue:43493", "right": "issue:43844", "accept": false, - "reason": "A SigLIP2 implementation mismatch is not the same as a DeepSpeed ZeRO-3 gradient explosion." + "reason": "SigLIP2 implementation discrepancy and DeepSpeed gradient inflation are unrelated." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "Embedding dtype drift in eval is unrelated to abnormal gradients during ZeRO-3 training." + "reason": "Embedding dtype drift and abnormal gradient growth under ZeRO-3 are different problems." }, { - "left": "issue:43646", - "right": "issue:44466", + "left": "issue:43278", + "right": "issue:43701", "accept": false, - "reason": "Custom model initialization breakage and lm_head.weight serialization inconsistency are different defects." + "reason": "A dtype change in evaluate is unrelated to resume_from_checkpoint key mismatch." }, { - "left": "issue:43278", - "right": "issue:43701", + "left": "issue:43646", + "right": "issue:44466", "accept": false, - "reason": "A dtype regression is not the same as a resume_from_checkpoint key mismatch." + "reason": "Custom model initialization breakage and lm_head.weight serialization inconsistency affect different code paths." }, { "left": "issue:46209", "right": "issue:46710", "accept": false, - "reason": "Both are regressions, but they affect different models and behaviors: FP8 inference on B200 vs accuracy drift after upgrade." + "reason": "FP8 inference on B200 and DeepSeek-R1 accuracy regression after upgrade are different issues." }, { "left": "issue:44568", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but one is mdeberta special-token behavior and the other is DeepSeek-Coder BPE marker/output corruption." + "reason": "Both are tokenizer regressions, but they hit different models with different special-token/BPE handling bugs." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "DiaConfig save/load token-ID loss is unrelated to MLukeTokenizer raising an AttributeError on tasks." + "reason": "DiaConfig save/load token ID loss and MLukeTokenizer AttributeError are unrelated." }, { "left": "issue:33357", "right": "issue:33666", "accept": false, - "reason": "A CLIP MacOS bus error and Qwen2-VL multi-GPU training are unrelated." + "reason": "A MacOS bus error in CLIP and Qwen2-VL multi-GPU training are unrelated." }, { "left": "issue:44849", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 hidden-states behavior is unrelated to the FP8 B200 inference issue." + "reason": "Qwen3.5 hidden-states behavior and FP8 inference on B200 are unrelated." }, { - "left": "issue:43572", + "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "Missing StableLmConfig pad_token_idx and an unreleased PEFT minimum version bump are unrelated." + "reason": "A janus model test failure and a PEFT version bump are not the same bug." }, { - "left": "issue:44792", + "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "A janus test failure does not match the PEFT version requirement issue." + "reason": "Qwen2-57B tensor-parallel OOM and PEFT version pinning are unrelated." }, { "left": "issue:44568", "right": "issue:45812", "accept": false, - "reason": "Different tokenizer regressions on different models; they are not the same concrete bug." + "reason": "Both are tokenizer regressions, but they affect different models and different tokenization rules; not the same underlying bug." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "Config field removal and unexpected init_empty_weights arguments are separate initialization problems." + "reason": "StableLmConfig pad_token_idx removal and init_empty_weights argument mismatch are different 5.x breakages." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "A specific model load failure is not the same as AutoTokenizer unconditionally calling model_info() and breaking offline mode." + "reason": "Specific model loading failure and HF_HUB_OFFLINE regression in _patch_mistral_regex are unrelated." }, { - "left": "issue:46821", - "right": "issue:46832", - "accept": false, - "reason": "TimesFM window_size AttributeError and RT-DETR forward shape mismatch are unrelated." - }, - { - "left": "issue:43334", - "right": "issue:45759", + "left": "issue:43824", + "right": "issue:44589", "accept": false, - "reason": "Both are Qwen multimodal config issues, but one is missing pad_token_id in Qwen3-VL text config while the other is from_config not unwrapping text_config for composite Qwen 3.5/3.6 configs." + "reason": "Importing Qwen2_5_VLForConditionalGeneration and missing Float8 storage object are unrelated errors." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage lookup failure and a flash_attn import_utils KeyError on Python 3.13 are different import/runtime issues." + "reason": "Float8 storage lookup failure and flash_attn KeyError on Python 3.13 are different runtime issues." }, { - "left": "issue:43824", - "right": "issue:44589", + "left": "issue:46821", + "right": "issue:46832", "accept": false, - "reason": "Missing Qwen2_5_VL import and float8 storage resolution failure are unrelated." + "reason": "TimesFM window_size AttributeError and RT-DETR forward TypeError affect different models and code paths." }, { - "left": "issue:44843", - "right": "issue:46489", + "left": "issue:44849", + "right": "issue:45357", "accept": false, - "reason": "Offline-mode model_info() behavior in tokenizer loading is unrelated to the DeepSeek-Coder tokenizer output regression." + "reason": "Qwen3.5 hidden-states bug and incorrect visual encoder keys on save_pretrained are distinct regressions." } ] }, @@ -8652,11 +8643,13 @@ "issue:43257", "issue:43299", "issue:43334", + "issue:43493", "issue:43504", "issue:43572", "issue:43606", "issue:43653", "issue:43696", + "issue:43701", "issue:43844", "issue:43873", "issue:43937", @@ -8664,7 +8657,6 @@ "issue:44112", "issue:44464", "issue:44493", - "issue:44554", "issue:44849", "issue:44869", "issue:44964", @@ -8673,11 +8665,10 @@ "issue:45200", "issue:45229", "issue:45245", - "issue:45357", - "issue:45446", "issue:45520", "issue:45584", "issue:45588", + "issue:45759", "issue:45823", "issue:46050", "issue:46285", @@ -8690,24 +8681,24 @@ "issue:47145" ], "soft_pairs": [ - "issue:44849|issue:45357", + "issue:43334|issue:45759", "issue:43504|issue:43606", "issue:43334|issue:43572", "issue:45245|issue:46710", - "issue:43257|issue:43299", "issue:44493|issue:46285", + "issue:43257|issue:43299", "issue:43504|issue:43653", "issue:43606|issue:44112", - "issue:45072|issue:46832", "issue:44849|issue:47145", + "issue:45072|issue:46832", "issue:46399|issue:46846", "issue:44464|issue:45584", "issue:43844|issue:44964", - "issue:43994|issue:44493", "issue:43873|issue:46050", + "issue:43994|issue:44493", "issue:29159|issue:32090", - "issue:44869|issue:46752", "issue:43937|issue:46701", + "issue:44869|issue:46752", "issue:45200|issue:45229", "issue:45229|issue:45823", "issue:42490|issue:43696", @@ -8716,198 +8707,198 @@ "issue:28218|issue:34567", "issue:29942|issue:30333", "issue:15354|issue:38175", - "issue:44554|issue:45446" + "issue:43493|issue:43701" ], - "prepared_review_unit_hash": "9cdd991647efa2e559d3e237908b5cb0e2081c7f04ce00300c6386566b230b3d", + "prepared_review_unit_hash": "b83ee30c7b4bbea5ddba94790b69c48940db45ca1282c3f25a48fd6b2be08b36", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35880, - "estimated_input_tokens": 8970, - "estimated_eval_tokens": 18196 + "serialized_chars": 35867, + "estimated_input_tokens": 8967, + "estimated_eval_tokens": 18190 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous Transformers bug reports with only superficial similarity by subsystem or model family. I do not see any true duplicate pair among the soft candidates.", + "summary": "This cluster is heterogeneous rather than a duplicate set. The soft edges are mostly title-level similarities across different models or subsystems, but they describe different failure modes and fixes, so all soft pairs are rejected.", "confidence": 0.96, - "canonical_issue_reason": "If one issue must anchor the cluster, issue 43334 is the clearest standalone bug report: a specific, reproducible loading failure with an exact AttributeError and narrow scope. That said, it does not appear to be a duplicate of the others.", - "canonical_pr_reason": null, - "best_issue_reason": "Issue 43334 is the best representative issue here because it is concrete, actionable, and well-scoped, unlike the rest of the set which spans unrelated model, tokenizer, loading, and training failures.", - "best_pr_reason": null, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated bugs in tokenizers, Qwen loading, CI, quantization, generation, and config handling.", + "canonical_pr_reason": "No PRs are present in this cluster.", + "best_issue_reason": "No single issue is a good global representative because the cluster does not share one underlying bug or change.", + "best_pr_reason": "No PRs are present in this cluster.", "soft_edge_verdicts": [ { - "left": "issue:44849", - "right": "issue:45357", + "left": "issue:43334", + "right": "issue:45759", "accept": false, - "reason": "Same model family, but different bugs: hidden-states output vs incorrect visual encoder keys on save_pretrained." + "reason": "Both involve Qwen multimodal loading, but one is a missing pad_token_id attribute in Qwen3-VL and the other is from_config not unwrapping text_config for Qwen 3.5/3.6; different root causes." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Unrelated failures: legacy pretrained-field loading vs CPU offload device mismatch." + "reason": "BEIT legacy-field loading failure and a Bark CPU-offload device mismatch are unrelated code paths and fixes." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Both involve missing config fields, but different models, different fields, and different failure paths." + "reason": "Both are config-attribute regressions after version updates, but they affect different models and different missing fields." }, { "left": "issue:45245", "right": "issue:46710", "accept": false, - "reason": "Completely different problems: category cardinality limit vs an accuracy regression in a specific model." + "reason": "A category-count runtime limit and a post-upgrade accuracy regression are not the same bug or fix." }, { - "left": "issue:43257", - "right": "issue:43299", + "left": "issue:44493", + "right": "issue:46285", "accept": false, - "reason": "Both are Qwen3 MOE-related, but one is accelerate/deepspeed weight conversion and the other is a dev-version loading breakage; not the same bug." + "reason": "Unexpected position-id keys and CLIP text-model structure inconsistency are different API/shape issues." }, { - "left": "issue:44493", - "right": "issue:46285", + "left": "issue:43257", + "right": "issue:43299", "accept": false, - "reason": "Different symptoms and code paths: unexpected position-id key vs CLIP text model structural inconsistency." + "reason": "Both mention Qwen MOE loading, but one is accelerate+deepspeed weight conversion and the other is a dev-version loading breakage for Qwen3VL; not the same concrete failure." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "Different subsystems entirely: semantic segmentation model loading vs tokenizer special-token registration." + "reason": "Legacy-field model loading and BigBirdTokenizer special-token registration are unrelated bugs." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "One is a real device-mismatch bug, the other is a CI test failure; no shared underlying defect." + "reason": "Both mention device-related CI failures, but they are for different models and different mismatch mechanisms." }, { - "left": "issue:45072", - "right": "issue:46832", + "left": "issue:44849", + "right": "issue:47145", "accept": false, - "reason": "Different models and failures: dtype mismatch in inference wrappers vs RT-DETR feature-level mismatch." + "reason": "Qwen3.5 hidden-states output bug and FP8 compressed-tensors weight_scale omission are different inference/serialization problems." }, { - "left": "issue:44849", - "right": "issue:47145", + "left": "issue:45072", + "right": "issue:46832", "accept": false, - "reason": "Different bug classes: output_hidden_states regression vs FP8 checkpoint loading without weight_scale." + "reason": "A dtype mismatch in inference and an RT-DETR feature-level mismatch in forward are unrelated code-path issues." }, { "left": "issue:46399", "right": "issue:46846", "accept": false, - "reason": "Both touch Qwen3.5, but one is a kernel registration crash and the other is a tensor-parallel plan omission." + "reason": "Missing decorator for apply_rotary_pos_emb registration and missing linear_attn entries in tp plan are different Qwen3.5 failures." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Different generation edge cases: compiled-forward inconsistency vs empty-transcription crash after token alignment." + "reason": "Chunked generation with compiled forward and Whisper empty-transcription failure after align_special_tokens are unrelated." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "Unrelated issues: ZeRO-3 gradient inflation vs a Phi-4 multimodal loading regression." + "reason": "Gradient growth with ZeRO-3 and a Phi-4 multimodal load failure are different bugs in different subsystems." }, { - "left": "issue:43994", - "right": "issue:44493", + "left": "issue:43873", + "right": "issue:46050", "accept": false, - "reason": "No shared bug: nonsensical SigLIP2 outputs vs unexpected position-id keys across many models." + "reason": "Quantization offloading not working and sinq save/load corruption are related broadly, but they are different concrete failure modes and fixes." }, { - "left": "issue:43873", - "right": "issue:46050", + "left": "issue:43994", + "right": "issue:44493", "accept": false, - "reason": "Related general area, but not the same concrete bug: offloading with quantization vs broken save/load for sinq-quantized models." + "reason": "SigLIP2 output mismatch and unexpected position-id keys are not the same underlying issue." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Tokenizer inconsistency and distributed broadcast NoneType errors are unrelated." + "reason": "Tokenizer slow/fast inconsistency and _gpu_broadcast_one rejecting NoneType are unrelated bugs." }, { - "left": "issue:44869", - "right": "issue:46752", + "left": "issue:43937", + "right": "issue:46701", "accept": false, - "reason": "Different crash sites and inputs: Whisper trailing-character timestamp decode vs empty chat-template input." + "reason": "Invalid GenerationConfig for GLM-5 and a DiffusionGemma pipeline failure are different problems." }, { - "left": "issue:43937", - "right": "issue:46701", + "left": "issue:44869", + "right": "issue:46752", "accept": false, - "reason": "Different behaviors and modules: invalid generation config vs broken image-text-to-text pipeline output type." + "reason": "Whisper timestamp decoding crash and apply_chat_template on empty list are distinct edge cases in different APIs." }, { "left": "issue:45200", "right": "issue:45229", "accept": false, - "reason": "Different Gemma 4 issues: missing mm_token_type_ids default vs multi-GPU CUDA OOM." + "reason": "Defaulting mm_token_type_ids and a multi-GPU CUDA OOM are not the same bug." }, { "left": "issue:45229", "right": "issue:45823", "accept": false, - "reason": "Same model family but different underlying problems: OOM vs device-map forward mismatch." + "reason": "Both concern Gemma4 inference, but one is OOM and the other is a device mismatch in PLE forward; different fixes." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "No overlap beyond model loading/runtime context: trailing-slash import warning vs plain CUDA OOM." + "reason": "Import warning for trailing slash paths and GPT-oss-20b CUDA OOM are unrelated." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Different failure modes: compile-time template-node restriction vs flash_attn import key error on Python 3.13." + "reason": "Compilation of non-template nodes and a Python 3.13 flash_attn import KeyError are different issues." }, { "left": "issue:45588", "right": "issue:46050", "accept": false, - "reason": "Different code paths: flash_attention sink-less model crash vs quantized checkpoint save/load corruption." + "reason": "Flash-attention integration crashing on s_aux=None and quantized sinq save/load breakage are unrelated." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Tokenizer spacing after added token is unrelated to TrainerState token-count tracking." + "reason": "Tokenizer spacing after added tokens and TrainerState token-count tracking are different tokenizer/training-state bugs." }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "One is a Flash Attention CI failure, the other is MLflow reporting on failed training jobs; not the same bug." + "reason": "Failing Flash Attention 2 tests and MLFlow not marking failed jobs as finished are unrelated CI/tracking issues." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript GeneratorExp export failure and zero-probability SigLIP2 outputs are unrelated." + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are unrelated model/runtime bugs." }, { - "left": "issue:44554", - "right": "issue:45446", + "left": "issue:43493", + "right": "issue:43701", "accept": false, - "reason": "Different issues: MPS attention correctness vs an incorrect PyTorch version guard in flex_attention imports." + "reason": "SigLIP2 HF vs JAX discrepancy and resume_from_checkpoint key mismatch do not share the same code-path or fix." } ] }, @@ -8948,7 +8939,6 @@ "issue:43550", "issue:43572", "issue:43575", - "issue:43701", "issue:43824", "issue:43844", "issue:43873", @@ -8956,16 +8946,18 @@ "issue:44186", "issue:44423", "issue:44521", + "issue:44554", "issue:44610", "issue:44734", - "issue:44737", "issue:44977", "issue:44987", "issue:45072", "issue:45198", "issue:45200", "issue:45290", + "issue:45381", "issue:45406", + "issue:45446", "issue:45520", "issue:45588", "issue:45759", @@ -8973,19 +8965,19 @@ "issue:46121", "issue:46178", "issue:46192", + "issue:46209", "issue:46369", "issue:46701", "issue:46710", - "issue:46821", - "issue:46894" + "issue:46821" ], "soft_pairs": [ - "issue:43493|issue:43701", + "issue:44554|issue:45446", "issue:43065|issue:46369", "issue:41720|issue:46710", "issue:44977|issue:46701", - "issue:37428|issue:43957", "issue:44423|issue:44734", + "issue:37428|issue:43957", "issue:43295|issue:43450", "issue:44977|issue:46710", "issue:45072|issue:45198", @@ -8997,235 +8989,235 @@ "issue:44186|issue:46192", "issue:43295|issue:44610", "issue:45823|issue:46192", - "issue:43572|issue:44987", "issue:43550|issue:43844", + "issue:43572|issue:44987", "issue:43450|issue:43575", "issue:29870|issue:29942", - "issue:37428|issue:43824", "issue:45520|issue:46178", + "issue:37428|issue:43824", "issue:45200|issue:45823", "issue:42490|issue:43493", - "issue:43295|issue:43493", "issue:39692|issue:46192", + "issue:43295|issue:43493", "issue:42915|issue:43824", "issue:44521|issue:45290", "issue:28218|issue:35141", - "issue:44737|issue:46894" + "issue:45381|issue:46209" ], - "prepared_review_unit_hash": "8e438f4af5a835c644cd0cfacca1d5f3e82e33f4ed17f161a11fe0bdd9f01693", + "prepared_review_unit_hash": "5334fc1cf7c254a03a5e32fac7f5689bc47acd7aba334f2d9eef5679251ca517", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36579, - "estimated_input_tokens": 9145, - "estimated_eval_tokens": 18546 + "serialized_chars": 37152, + "estimated_input_tokens": 9288, + "estimated_eval_tokens": 18832 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a true duplicate set: it mixes many unrelated Hugging Face bugs across tokenizers, multimodal processors, serving, flash-attention, device mapping, and config/loading regressions. None of the soft pairs look like the same underlying fix.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "None of the proposed soft pairs look like true duplicates; they span different models, subsystems, and failure modes with only superficial topic overlap.", + "confidence": 0.98, + "canonical_issue_reason": "No single issue is a clear canonical duplicate target here; the items are mostly unrelated bugs rather than one underlying problem.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No issue stands out as a global representative because the cluster does not converge on one concrete bug or code path.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43493", - "right": "issue:43701", + "left": "issue:44554", + "right": "issue:45446", "accept": false, - "reason": "Different problems: SigLIP2 implementation discrepancy vs a resume_from_checkpoint key mismatch. No shared code-path." + "reason": "Different problems: MPS attention correctness vs a PyTorch version-check/import issue in flex_attention." }, { "left": "issue:43065", "right": "issue:46369", "accept": false, - "reason": "Sam3PixelDecoder dummy Conv2d vs sam3-litetext memory overflow; unrelated symptoms and fixes." + "reason": "Both mention Sam3, but one is about a dummy Conv2d in the pixel decoder and the other is a memory overflow in sam3-litetext." }, { "left": "issue:41720", "right": "issue:46710", "accept": false, - "reason": "Both mention Qwen3/DeepSeek-style regression contexts, but one is a device-map cuda assert and the other is an accuracy regression after upgrade. Not the same bug." + "reason": "One is a CUDA assert during auto device mapping; the other is an accuracy regression after a version upgrade." }, { "left": "issue:44977", "right": "issue:46701", "accept": false, - "reason": "Qwen3.5 flash-attention generation failure is not the same as a diffusion pipeline output-type issue." + "reason": "Different models and failures: Qwen3.5 flash-attention generation vs a DiffusionGemma image-text pipeline issue." }, { - "left": "issue:37428", - "right": "issue:43957", + "left": "issue:44423", + "right": "issue:44734", "accept": false, - "reason": "Both are loading/import regressions, but they affect different symbols and different model-loading paths." + "reason": "Both are serve-related, but they hit different code paths: continuous batching multimodal crash vs KV-cache continuation indexing." }, { - "left": "issue:44423", - "right": "issue:44734", + "left": "issue:37428", + "right": "issue:43957", "accept": false, - "reason": "Serve crashes in different places: multimodal continuous batching vs KV-cache continuation tensor indexing." + "reason": "An import error for a flash-attention helper is not the same as meta-device model loading breaking some models." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Regression around processor.tokenizer/images-to-tokenizer access is distinct from batched video shape handling." + "reason": "Different regressions: processor/tokenizer access in custom code vs batched video processor output shape." }, { "left": "issue:44977", "right": "issue:46710", "accept": false, - "reason": "Generation broken under flash-attention vs a broad accuracy regression after version upgrade; not mergeable as one fix." + "reason": "Unrelated: flash-attention generation bug in Qwen3.5 vs an accuracy regression in DeepSeek." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "Different dtype/tokenization failures in different model families; only broad training/inference bug category overlaps." + "reason": "Different subsystems and symptoms: dtype mismatch in inference vs save_pretrained/tokenization failure." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention sink-less handling and AutoModelForCausalLM text_config unwrapping are separate code paths." + "reason": "Different bugs: flash_attention integration crash on s_aux=None vs not unwrapping text_config for composite multimodal configs." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to rope validation type handling." + "reason": "Gemma4 serve processor attribute error is unrelated to rope-parameter validation of JSON-loaded lists." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Video-input ValueError in Qwen3OmniMoe vs Gemma 4 mm_token_type_ids defaulting; different processing logic." + "reason": "Video processing in Qwen3OmniMoe is unrelated to Gemma4 mm_token_type_ids defaults." }, { "left": "issue:43572", "right": "issue:46821", "accept": false, - "reason": "StableLmConfig pad_token_idx regression and TimesFM window_size AttributeError are unrelated config bugs." + "reason": "StableLm pad_token_idx config regression and TimesFM window_size attribute error are distinct." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor/tokenizer regression is not the same as quantization offloading behavior." + "reason": "Processor/tokenizer regression is unrelated to quantization offloading behavior." }, { "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes on NER/padding is unrelated to return_tensors being ignored with text_kwargs." + "reason": "Tokenizer crash on NER/padding is not the same as return_tensors being ignored when text_kwargs is passed." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "Custom model processor/tokenizer regression and OmDet-Turbo image-size mismatch are different model-specific issues." + "reason": "Different issues: processor.tokenizer regression vs OmDet-Turbo processor/model input-size mismatch." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device mismatch during forward is unrelated to tensor-return behavior in processing." + "reason": "Gemma4 device-mismatch during forward is unrelated to return_tensors handling." }, { - "left": "issue:43572", - "right": "issue:44987", + "left": "issue:43550", + "right": "issue:43844", "accept": false, - "reason": "StableLmConfig pad_token_idx vs loading physical-intelligence/fast under transformers>=5.1.0 are different failures." + "reason": "One is a torch.compile/SDPA bug; the other is a training gradient issue with ZeRO-3." }, { - "left": "issue:43550", - "right": "issue:43844", + "left": "issue:43572", + "right": "issue:44987", "accept": false, - "reason": "torch.compile + SDPA on Bamba-9B-v2 is a model/backend issue, not the same as ZeRO-3 gradient growth." + "reason": "StableLm config padding regression and a physical-intelligence loading failure are separate bugs." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Batched video shape handling and tensor-parallel OOM when loading Qwen2-57B are not the same defect." + "reason": "Batched video shape bug and tensor-parallel OOM are unrelated failures." }, { "left": "issue:29870", "right": "issue:29942", "accept": false, - "reason": "Bart generate decoder_attention_mask bug and Flash Attention 2 test failures share generation-related context only, not the same code-path." + "reason": "BART decoder_attention_mask UnboundLocalError is not the same as Flash Attention 2 test failures." }, { - "left": "issue:37428", - "right": "issue:43824", + "left": "issue:45520", + "right": "issue:46178", "accept": false, - "reason": "Different missing-import regressions: one for _flash_supports_window_size, one for Qwen2_5_VLForConditionalGeneration." + "reason": "Both involve import/backends mapping, but they concern different backends and different error conditions." }, { - "left": "issue:45520", - "right": "issue:46178", + "left": "issue:37428", + "right": "issue:43824", "accept": false, - "reason": "Python 3.13 flash_attn import KeyError and tensorflow_text backend mapping error are unrelated import-system issues." + "reason": "Different ImportErrors for different missing symbols/modules." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Both involve Gemma 4, but one is missing mm_token_type_ids defaults and the other is a device mismatch in PLE forward." + "reason": "Same model family, but one is a missing default token-type tensor and the other is a forward device placement bug." }, { "left": "issue:42490", "right": "issue:43493", "accept": false, - "reason": "Trailing-slash local-model import warning is unrelated to SigLIP2 implementation mismatch." + "reason": "Local-path import warning is unrelated to the SigLIP2 implementation discrepancy." }, { - "left": "issue:43295", - "right": "issue:43493", + "left": "issue:39692", + "right": "issue:46192", "accept": false, - "reason": "Different multimodal regression classes: processor/tokenizer access vs HF/JAX implementation discrepancy." + "reason": "Documentation/example errors for SigLIP2 are not the same as return_tensors being ignored." }, { - "left": "issue:39692", - "right": "issue:46192", + "left": "issue:43295", + "right": "issue:43493", "accept": false, - "reason": "SigLIP2 doc example errors are documentation/example failures, not the same as return_tensors being ignored." + "reason": "Processor/tokenizer regression and SigLIP2 parity discrepancy are different bugs." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and Qwen2_5_VL import error are separate model/load issues." + "reason": "FineGrainedFP8Config failure in Qwen3Moe is unrelated to the Qwen2.5-VL import error." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "assistant_masks all-zero for multimodal inputs is distinct from tokenize=True crashing on tool-call assistant messages." + "reason": "Both involve apply_chat_template, but one is wrong assistant masks for multimodal inputs and the other is a crash on tool-call messages with no content." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Tokenizer extra-space bug and output embedding reinitialization after resize are completely different behaviors." + "reason": "Tokenizer spacing after added tokens and embedding reinitialization after resize are different bugs." }, { - "left": "issue:44737", - "right": "issue:46894", + "left": "issue:45381", + "right": "issue:46209", "accept": false, - "reason": "XLNet CPU arange regression and CohereASR double-shift training-loss bug affect different models and code paths." + "reason": "Qwen2.5-VL video position IDs and FP8 inference on B200 are unrelated." } ] }, @@ -9276,6 +9268,7 @@ "issue:44466", "issue:44561", "issue:44610", + "issue:44737", "issue:44792", "issue:44857", "issue:44898", @@ -9284,7 +9277,6 @@ "issue:45198", "issue:45200", "issue:45325", - "issue:45381", "issue:45412", "issue:45563", "issue:45593", @@ -9293,17 +9285,17 @@ "issue:46234", "issue:46285", "issue:46512", - "issue:46821", + "issue:46894", "issue:46899", "issue:46932" ], "soft_pairs": [ "issue:44466|issue:45127", - "issue:45381|issue:46209", + "issue:44737|issue:46894", "issue:43278|issue:43856", - "issue:43452|issue:46512", "issue:43653|issue:44112", "issue:43824|issue:43975", + "issue:43452|issue:46512", "issue:43262|issue:43450", "issue:43950|issue:46285", "issue:43334|issue:43525", @@ -9321,215 +9313,208 @@ "issue:44610|issue:44857", "issue:43819|issue:45083", "issue:43827|issue:43901", - "issue:44857|issue:45593", "issue:43606|issue:45083", - "issue:37428|issue:43425", + "issue:44857|issue:45593", "issue:43388|issue:43526", - "issue:46512|issue:46821", + "issue:37428|issue:43425", "issue:43335|issue:43653" ], - "prepared_review_unit_hash": "b40ba0aeaf74b18fb39c6eb1cf79e478520e30ff43be06feea42d9e30472b4ad", + "prepared_review_unit_hash": "cda73577c37f843c8d9b96bfa411b06e9b3519c4a421c14c28ea72e1c25aec14", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 36225, - "estimated_input_tokens": 9057, - "estimated_eval_tokens": 18370 + "soft_pair_count": 28, + "serialized_chars": 35952, + "estimated_input_tokens": 8988, + "estimated_eval_tokens": 18232 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose semantic grouping of mostly unrelated Transformers bug reports and regressions. The soft pairs share superficial themes (missing attrs, processor bugs, v5 regressions) but do not look like the same underlying change, so none should be merged as duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 43950 is the best representative anchor: it is a broad, cross-cutting v5 regression in core model loading/serialization, with clearer general impact than the more model-specific reports.", + "summary": "The cluster is heterogeneous: the items span unrelated regressions in model configs, tokenizers, processors, training, docs, and CI. None of the soft pairs describe the same concrete bug or change, so all soft edges are rejected and there is no clear canonical issue/PR for the cluster.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43950 is the strongest global issue candidate because it targets a core loader path, has a concrete repro, and affects multiple models rather than a single model family.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Both touch tied weights / serialization, but one is about device-dependent lm_head serialization and the other about LoRA merge collapse; different bugs and fixes." + "reason": "Both concern tied-weight behavior, but one is about device-dependent serialization of lm_head.weight and the other about LoRA merge collapse with extended vocab; different failure modes and fixes." }, { - "left": "issue:45381", - "right": "issue:46209", + "left": "issue:44737", + "right": "issue:46894", "accept": false, - "reason": "Qwen2.5-VL video position IDs vs FP8 inference failure on B200 are unrelated code paths and symptoms." + "reason": "XLNet CPU arange placement is unrelated to CohereASR's training-loss double-shift bug; different models, code paths, and symptoms." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Training/eval dtype mismatch is not the same as Qwen3 MoE memory inefficiency." - }, - { - "left": "issue:43452", - "right": "issue:46512", - "accept": false, - "reason": "gguf_file loading breakage and stale config.dtype after from_config are separate loader API regressions." + "reason": "BF16\u2192FP32 dtype drift in evaluate is not the same as Qwen3 MoE memory inefficiency during training." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "Tokenizer special-token registration and a GraniteSpeech CI device-override test failure are unrelated." + "reason": "A tokenizer special-token registration bug is unrelated to a GraniteSpeech CI device-override test regression." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Importing a missing Qwen2.5-VL class and incorrect detokenization in DeepSeek-Coder are different failures." + "reason": "Missing Qwen2_5_VL import/export is a namespace/API break; incorrect detokenization in deepseek-coder is a separate tokenizer behavior bug." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file/from_pretrained loading issues are unrelated to from_config(dtype=...) leaving config.dtype stale." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate handling and batched video output shape are distinct processor bugs." + "reason": "Audio chat-template sampling-rate defaults and batched video output shapes are different processor bugs." }, { "left": "issue:43950", "right": "issue:46285", "accept": false, - "reason": "Non-persistent buffer corruption in from_pretrained and CLIP text-model structure differences do not share the same fix." + "reason": "Non-persistent buffer corruption during from_pretrained is unrelated to CLIP text-model structural inconsistency." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Both mention pad_token_id, but they affect different config classes/models and likely require separate fixes." + "reason": "Both mention pad_token_id, but they are separate missing-attribute bugs in different model configs (Qwen3-VL vs Llama4)." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "FastSpeech2Conformer tokenizer loading and Wav2Vec2 save/tokenization failures are different tokenizer/model issues." + "reason": "FastSpeech2ConformerTokenizer loading failure is unrelated to Wav2Vec2 save_pretrained/tokenization failure." }, { "left": "issue:43526", "right": "issue:46192", "accept": false, - "reason": "reduce_labels dropping labels and return_tensors being ignored are separate preprocessing bugs." + "reason": "BeitImageProcessorFast label reduction and text_kwargs/return_tensors handling are distinct preprocessing issues." }, { "left": "issue:45200", "right": "issue:46899", "accept": false, - "reason": "mm_token_type_ids defaulting to zeros and unsafe dtype casting in quantized multimodal embedders are different Gemma4 failures." + "reason": "Defaulting mm_token_type_ids to zeros for text-only fine-tuning is unrelated to unsafe dtype casting in quantized multimodal embedders." }, { "left": "issue:45200", "right": "issue:46932", "accept": false, - "reason": "The titles mention Gemma4 multimodal behavior, but one is about token-type IDs and the other about quantized embedders/dtype casting." + "reason": "These are different Gemma4 bugs: one is text-only mm_token_type_ids defaults, the other is quantized vision/audio embedder casting." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "A janus generate_images test failure and a stale num_return_sequences warning are not the same bug." + "reason": "A failed janus image-generation test is not the same as a stale paged-generate warning for num_return_sequences." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Perceiver resolution handling and RT-DETR memory release are unrelated issues." + "reason": "Perceiver non-default resolution handling and RT-DETR memory release are unrelated model issues." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent conversion mismatch and GraniteSpeech CI device-override flakiness have no shared concrete code path." + "reason": "DAC.from_latents mismatch with missing STE is unrelated to a GraniteSpeech CI test failure." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "SwitchTransformers sparse-layer config creation and Bamba SDPA torch.compile failure are separate model/runtime bugs." + "reason": "SwitchTransformers sparse-layer creation and Bamba SDPA torch.compile failure affect different models and code paths." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Gemma4 mm_token_type_ids defaulting and Qwen2.5-VL still-image rope indexing are different multimodal processing problems." + "reason": "Gemma4 text-only mm_token_type_ids defaults are unrelated to Qwen2.5-VL rope index scaling for still images." }, { "left": "issue:44561", "right": "issue:46234", "accept": false, - "reason": "Removing is_torch_fx_available and missing is_falcon_mamba_ssm_available are different import-utility breakages." + "reason": "These are separate import_utils API removals/additions affecting different helpers and remote-code models." }, { "left": "issue:41720", "right": "issue:46209", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and FP8 inference on B200 are unrelated hardware/runtime failures." + "reason": "Qwen3 auto device-mapping cuda asserts on A800 and FP8 inference failure on B200 are different hardware/model issues." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "OmDet input-size mismatch and LwDetrImageLoss AMP crash are distinct model/precision bugs." + "reason": "OmDet-Turbo processor size mismatch and LwDetrImageLoss AMP crash are unrelated vision bugs." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "DAC from_latents mismatch and qwen3_omni_moe feature-length helper behavior are unrelated." + "reason": "DAC latent reconstruction mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs regressions, but they concern different pipeline pages and different deprecated arguments." - }, - { - "left": "issue:44857", - "right": "issue:45593", - "accept": false, - "reason": "AMP/CUDA crash in LwDetrImageLoss and missing auxiliary losses in D-FINE are different training-path problems." + "reason": "Both are docs regressions, but they affect different pipeline pages and different deprecations; not the same change." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "CPU offload device mismatch in bark-small and qwen3_omni_moe feature-length helper behavior are unrelated." + "reason": "CPU offload device mismatch in bark-small is unrelated to qwen3_omni_moe feature-length calculation." }, { - "left": "issue:37428", - "right": "issue:43425", + "left": "issue:44857", + "right": "issue:45593", "accept": false, - "reason": "A missing flash-attention helper import and torch 2.10 incompatibility are not the same concrete regression." + "reason": "AMP crash in LwDetrImageLoss is not the same underlying issue as D-FINE omitting auxiliary losses when denoising is off." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "gather_for_metrics truncating tuple labels and BeitImageProcessorFast reducing labels differently are separate label-handling bugs." + "reason": "Mask2Former gather_for_metrics label truncation is unrelated to BeitImageProcessorFast reduce_labels behavior." }, { - "left": "issue:46512", - "right": "issue:46821", + "left": "issue:37428", + "right": "issue:43425", "accept": false, - "reason": "Stale config.dtype after from_config and TimesFM window_size attribute errors are unrelated API issues." + "reason": "A missing flash-attention helper import and broad Torch 2.10 incompatibility are different compatibility problems." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "Sparse-layer creation in SwitchTransformers and BigBird tokenizer special-token registration are unrelated subsystems." + "reason": "Sparse-layer misconfiguration in SwitchTransformers and BigBirdTokenizer special-token registration are unrelated bugs." } ] }, @@ -9598,19 +9583,20 @@ "issue:46285", "issue:46326", "issue:46424", - "issue:46431", - "issue:46566", + "issue:46512", + "issue:46821", "issue:46858" ], "soft_pairs": [ "issue:44206|issue:45584", + "issue:46512|issue:46821", "issue:45759|issue:46285", "issue:46424|issue:46858", "issue:29942|issue:32090", "issue:38175|issue:45412", "issue:28282|issue:33357", - "issue:43611|issue:44466", "issue:43335|issue:43504", + "issue:43611|issue:44466", "issue:44855|issue:45083", "issue:43550|issue:44112", "issue:45520|issue:46170", @@ -9621,206 +9607,205 @@ "issue:43335|issue:45593", "issue:43278|issue:43825", "issue:34689|issue:35141", - "issue:45588|issue:45859", "issue:42490|issue:43854", "issue:45290|issue:46326", + "issue:45588|issue:45859", "issue:29127|issue:32090", "issue:44514|issue:44521", "issue:39692|issue:43526", "issue:43606|issue:44857", - "issue:43819|issue:44493", - "issue:46431|issue:46566" + "issue:43819|issue:44493" ], - "prepared_review_unit_hash": "101b96518e43af6c00b41597898c9e23642c191f372816513b17f66bbb937777", + "prepared_review_unit_hash": "32778769ea29b598b25fb410eb03766298a700194397e92afeea9901a23a1b36", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36256, - "estimated_input_tokens": 9064, - "estimated_eval_tokens": 18384 + "serialized_chars": 36239, + "estimated_input_tokens": 9060, + "estimated_eval_tokens": 18376 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous: most items are unrelated regressions across different models, utilities, and backends. I do not see any soft pair that clearly represents the same underlying bug/change, so none should be merged as duplicates.", - "confidence": 0.86, - "canonical_issue_reason": "issue:44521 is the most central active issue in the list: it targets the broadly used multimodal `apply_chat_template` path, has multiple inbound references, and is more representative of the cluster than the narrower one-off bugs.", + "summary": "This cluster is broad and mostly heterogeneous: many distinct Transformers bugs around loading/configuration, chat templates, cache behavior, and backend-specific failures. There is no clear duplicate pair among the soft candidates, so the cluster should be treated as noisy rather than a single underlying defect.", + "confidence": 0.34, + "canonical_issue_reason": "issue:43611 is the broadest, most cross-cutting regression in the set (model loading with base_model_prefix in v5) and is a better representative than the many narrowly scoped model-specific or backend-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the best representative issue because it is current, has the broadest user impact among the listed issues, and sits in the most recurring theme here (chat-template handling).", + "best_issue_reason": "issue:43611 is the best single issue representative because it describes a general loading regression affecting multiple models, which is closer to the cluster\u2019s load/config-heavy center of mass than the very specific edge cases elsewhere.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "Different code paths: one is a `LasrFeatureExtractor` argument regression, the other is Whisper generation on empty transcription after token alignment." + "reason": "Both are runtime crashes, but one is a feature-extractor argument issue and the other is an empty-transcription generation bug; different code paths and fixes." + }, + { + "left": "issue:46512", + "right": "issue:46821", + "accept": false, + "reason": "Stale dtype metadata in from_config and a TimesFM window_size AttributeError are unrelated bugs despite both being configuration-facing." }, { "left": "issue:45759", "right": "issue:46285", "accept": false, - "reason": "Both mention model/config structure, but one is `AutoModelForCausalLM.from_config` unwrapping `text_config` and the other is a CLIP class structure inconsistency." + "reason": "One is about AutoModelForCausalLM unwrapping text_config for composite multimodal configs; the other is a CLIP class structure inconsistency. Not the same defect." }, { "left": "issue:46424", "right": "issue:46858", "accept": false, - "reason": "Both involve cache behavior, but one is `max_cache_length` semantics for Quantized Cache and the other is `get_seq_length()` returning a device tensor breaking static-cache arithmetic." + "reason": "Both involve cache mechanics, but one is about max_cache_length handling and the other is a device-tensor seq_length arithmetic bug in static cache; distinct failures." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Unrelated failures: Flash Attention 2 test breakage versus Trainer GPU broadcast rejecting `NoneType`." + "reason": "Flash Attention test failures and a Trainer _gpu_broadcast_one NoneType error are unrelated issues in different subsystems." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "Different subsystems and symptoms: SigLIP2 zero probabilities versus RT-DETR memory not being released on deletion." + "reason": "Unexpected zero probabilities and an RT-DETR memory leak are completely different model/runtime problems." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "One is a missing PyTorch import error, the other is a MacOS bus error loading a CLIP model; not the same bug." + "reason": "Missing PyTorch ImportError and a MacOS bus error with a community CLIP model are not the same bug." }, { - "left": "issue:43611", - "right": "issue:44466", + "left": "issue:43335", + "right": "issue:43504", "accept": false, - "reason": "Both are v5/model-loading related, but one is about `base_model_prefix` loading and the other about device-dependent `lm_head.weight` serialization." + "reason": "SwitchTransformers sparse-layer creation and BEiT legacy-field loading are different model-specific configuration bugs." }, { - "left": "issue:43335", - "right": "issue:43504", + "left": "issue:43611", + "right": "issue:44466", "accept": false, - "reason": "Different models and failure modes: SwitchTransformers sparse-layer creation versus BEiT legacy-field loading." + "reason": "Model-loading failure via base_model_prefix is not the same as inconsistent lm_head.weight serialization depending on device." }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "`deberta` import/parsing issue on Python 3.13 is unrelated to the Qwen Omni feature-extraction helper behavior." + "reason": "Python 3.13 IndentationError in DebertaV2Model and a qwen3_omni_moe helper output-length bug are unrelated." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "A torch.compile/SDPA model bug is not the same as a GraniteSpeech CI test with stale device override state." + "reason": "torch.compile/SDPA failure for Bamba-9B-v2 and a stale-device override CI failure are different issues." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Different import/platform problems: `flash_attn` lookup on Python 3.13 versus Windows ROCm/torch.distributed dependency issues." + "reason": "A flash_attn import key error on Python 3.13 and a Windows ROCm torch.distributed import/type error are separate platform-specific problems." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they involve different symbols and modules (`_flash_supports_window_size` vs `HybridCache`)." + "reason": "Cannot import _flash_supports_window_size and cannot import HybridCache are both import errors, but they stem from different symbols and modules." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers config bug and Qwen Omni output-length helper bug are unrelated." + "reason": "A sparse-layer creation bug and a helper function returning wrong feature lengths are unrelated." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Pegasus v4/v5 output differences are not the same as Qwen3.5 recurrent-state reset with cache." + "reason": "Pegasus output differences across versions and qwen3 recurrent state resets are different model behaviors with different fixes." }, { "left": "issue:43335", "right": "issue:44112", "accept": false, - "reason": "A model-construction bug and a CI test flake on a different model are not the same issue." + "reason": "SwitchTransformersConfig sparse-layer initialization and GraniteSpeech stale device override are not the same code-path problem." }, { "left": "issue:43335", "right": "issue:45593", "accept": false, - "reason": "SwitchTransformers sparse-layer creation is unrelated to D-FINE auxiliary-loss behavior." + "reason": "Sparse encoder-layer creation and D-FINE auxiliary-loss handling when denoising is off are unrelated model-specific bugs." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Different API surfaces: evaluation dtype drift versus a pipeline error message about translation task support." + "reason": "Embedding dtype staleness during evaluate and a pipeline task-support warning bug do not share the same underlying cause." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Both concern model loading/weights, but one is a specific Llama vision regression and the other is token-embedding resizing reinitializing output embeddings." - }, - { - "left": "issue:45588", - "right": "issue:45859", - "accept": false, - "reason": "Both touch MoE/flash/cpu-offload areas, but one is an `AttributeError` on `s_aux=None` and the other is a rotary-embedding CPU-offload incompatibility." + "reason": "Llama 3.2 Vision model loading regression and token-embedding resize/post_init reinitialization are different issues." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "A trailing-slash local-path import warning is unrelated to loading a specific remote model in unit tests." + "reason": "Trailing-slash local model import warning and GLM-4.7-Flash test-loading failures are distinct loading problems." }, { "left": "issue:45290", "right": "issue:46326", "accept": false, - "reason": "Different chat-template bugs: one crashes on assistant tool-call messages with no content, the other is `continue_final_message` for `mistral_common` tokenizers." + "reason": "Both concern apply_chat_template, but one crashes on assistant tool-call messages with no content while the other is a mistral_common continue_final_message bug; not the same defect." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "A flash_attention integration crash on s_aux=None and a Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are different backend issues." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "LayoutLMv3 error-message clarity is unrelated to Trainer GPU broadcast failing on `NoneType`." + "reason": "LayoutLMv3 error-message clarity and Trainer GPU broadcast NoneType support are unrelated." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both involve `apply_chat_template` for multimodal/Qwen-VL paths, but one is a batched-input crash with `padding=False` while the other is all-zero assistant masks; different failure modes." + "reason": "Both are Qwen2_5_VL apply_chat_template problems, but one is a batching/padding crash and the other is all-zero assistant masks; similar area, different bug." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "SigLIP2 documentation example errors are unrelated to BEiT fast image processor label reduction." + "reason": "SigLIP2 docs-example errors and BEiT fast image-processor reduce_labels behavior are unrelated." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "CPU offload device mismatch in Bark is not the same bug as LwDetrImageLoss crashing under float16 AMP on CUDA." + "reason": "CPU-offload device mismatch in bark-small and float16 AMP crash in LwDetrImageLoss are different hardware/backend bugs." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC latent/forward mismatch and unexpected position-id keys are different model behaviors." - }, - { - "left": "issue:46431", - "right": "issue:46566", - "accept": false, - "reason": "Missing chat template on a processor and DiffusionGemma remaining on `meta` under `device_map=\"auto\"` are unrelated." + "reason": "DAC.from_latents/STE mismatch and unexpected position-id keys are unrelated model internals." } ] }, @@ -9853,7 +9838,6 @@ "issue:29942", "issue:31501", "issue:32090", - "issue:33290", "issue:33357", "issue:34689", "issue:35141", @@ -9876,7 +9860,6 @@ "issue:44617", "issue:44779", "issue:44857", - "issue:44977", "issue:45072", "issue:45083", "issue:45127", @@ -9890,237 +9873,232 @@ "issue:45823", "issue:46234", "issue:46424", + "issue:46431", + "issue:46566", "issue:46697", "issue:47000" ], "soft_pairs": [ - "issue:43720|issue:44112", + "issue:46431|issue:46566", "issue:28218|issue:32090", "issue:45310|issue:45478", + "issue:43720|issue:44112", "issue:39692|issue:44493", "issue:44448|issue:44779", - "issue:44206|issue:45823", "issue:43701|issue:45083", + "issue:44206|issue:45823", "issue:44617|issue:46697", "issue:26413|issue:31501", "issue:45127|issue:45245", "issue:45459|issue:46234", - "issue:44857|issue:45072", "issue:32090|issue:33357", "issue:42490|issue:43575", + "issue:44857|issue:45072", "issue:43550|issue:45083", "issue:45593|issue:47000", - "issue:27957|issue:34689", "issue:44464|issue:46424", + "issue:27957|issue:34689", "issue:38175|issue:44493", "issue:29942|issue:35141", "issue:29870|issue:32090", "issue:10000|issue:15354", "issue:43278|issue:43329", - "issue:37428|issue:42175", "issue:45290|issue:45381", - "issue:44448|issue:44977", - "issue:29870|issue:33290" + "issue:37428|issue:42175" ], - "prepared_review_unit_hash": "64e43fb628de6f4c2572366236f89c3992671e2f645a1c7e5184d459898b79be", + "prepared_review_unit_hash": "e054b112e26cfccf0b6b83a7f7d42abc39450234d2a6b3ce3d9e836fbbc4cb4b", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35795, - "estimated_input_tokens": 8949, - "estimated_eval_tokens": 18154 + "soft_pair_count": 26, + "serialized_chars": 35566, + "estimated_input_tokens": 8892, + "estimated_eval_tokens": 18040 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly noisy and heterogeneous; almost all candidate edges are false matches. The only clear duplicate pair is the two identical Qwen3.5 MoE `from_pretrained` regression reports.", + "summary": "Most pairs are unrelated and should stay separate. The only clear duplicate is the two Qwen3.5 Moe from_pretrained reports: issue:45310 and issue:45478.", "confidence": 0.91, - "canonical_issue_reason": "issue:45310 is the best canonical issue because issue:45478 is an apparent later duplicate with the same title and failure mode, while 45310 has the fuller original report.", + "canonical_issue_reason": "issue:45310 is the earlier report in the same Qwen3.5 Moe from_pretrained error family, with a matching title to issue:45478.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the strongest representative in this set: it is the only exact same-bug duplicate pair and has the more complete discussion, making it the best anchor for triage.", + "best_issue_reason": "issue:45310 is the strongest canonical issue candidate because it is the earlier report and clearly names the same Qwen3.5 Moe from_pretrained failure.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43720", - "right": "issue:44112", + "left": "issue:46431", + "right": "issue:46566", "accept": false, - "reason": "Different bugs: one is an Accelerate loading failure for packed BitNet weights; the other is a CI failure in a GraniteSpeech device override test." + "reason": "Both touch model-loading behavior, but one is about a missing chat template error and the other is a device_map/meta-device placement bug; different failures and code paths." }, { "left": "issue:28218", "right": "issue:32090", "accept": false, - "reason": "Unrelated: tokenizer spacing after added tokens versus a Trainer `_gpu_broadcast_one` NoneType type error." + "reason": "Tokenizer spacing after added tokens is unrelated to a Trainer NoneType broadcast error." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "Same underlying bug: identical Qwen3.5 MoE `from_pretrained` failure on transformers>=5.4.0; `45478` looks like a duplicate report of `45310`." + "reason": "Same model family and same from_pretrained error theme, with matching titles; issue:45310 is the earlier report." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "Packed-weight loading in accelerate and a GraniteSpeech CI device-override test are different bugs." }, { "left": "issue:39692", "right": "issue:44493", "accept": false, - "reason": "Different problems: SigLIP2 docs/model-processor/quantization issues versus unexpected position-id key behavior across many models." + "reason": "SigLIP2 docs/model-processor mismatch and unexpected position-id keys are not the same underlying issue." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Both are version-regression/output complaints, but they affect different models and different code paths (Pegasus summarization vs Deepseek tokenizer)." + "reason": "Pegasus v4/v5 output divergence and Deepseek tokenizer regression are different code paths." }, { - "left": "issue:44206", - "right": "issue:45823", + "left": "issue:43701", + "right": "issue:45083", "accept": false, - "reason": "Different subsystems: a LaserFeatureExtractor `center` argument regression versus a Gemma4 PLE device-map mismatch during forward." + "reason": "resume_from_checkpoint key mismatch is unrelated to qwen3_omni_moe feature-length behavior." }, { - "left": "issue:43701", - "right": "issue:45083", + "left": "issue:44206", + "right": "issue:45823", "accept": false, - "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen3 Omni MoE feature-length helper behavior." + "reason": "Unsupported center argument in LasrFeatureExtractor and Gemma4 device mismatch are distinct bugs." }, { "left": "issue:44617", "right": "issue:46697", "accept": false, - "reason": "Different errors in different models: Sam3Video CUDA OOM versus Swin2SR over-padding in the image processor." + "reason": "Sam3Video CUDA OOM and Swin2SR over-padding are unrelated." }, { "left": "issue:26413", "right": "issue:31501", "accept": false, - "reason": "Both mention training startup/iterator behavior, but they are distinct failure modes and not the same bug." + "reason": "Both relate to training startup, but the concrete failure modes are different." }, { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "Different issues: LoRA merge with extended vocabulary causing collapse versus a categories-limit runtime error." + "reason": "LoRA merge output collapse and category-count overflow are separate problems." }, { "left": "issue:45459", "right": "issue:46234", "accept": false, - "reason": "Tokenizer protobuf error masking is unrelated to a missing import utility function for Falcon Mamba availability." - }, - { - "left": "issue:44857", - "right": "issue:45072", - "accept": false, - "reason": "Both are CI/model dtype-related, but one is LwDetr loss AMP failure and the other is SwitchTransformers/TimmWrapper dtype mismatch; not the same bug." + "reason": "Tokenizer protobuf error handling and a missing import_utils backend symbol are unrelated." }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Different failures: Trainer broadcast NoneType error versus MacOS bus error with a community CLIP model." + "reason": "Trainer broadcast NoneType and MacOS CLIP bus error are different runtime failures." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "Import warning for local paths ending with a slash is unrelated to tensor-parallel OOM when loading Qwen2-57B-A14B-Instruct." + "reason": "Local-path import warning and TP OOM are not the same bug." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "AMP/CUDA loss crash and bfloat16 dtype mismatch in CI are different issues." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "Different code paths: torch.compile/SDPA failure in Bamba-9B-v2 versus a feature-length helper bug in qwen3_omni_moe." + "reason": "torch.compile + SDPA failure and feature-length helper behavior are unrelated." }, { "left": "issue:45593", "right": "issue:47000", "accept": false, - "reason": "Different detection-loss bugs: missing auxiliary losses when denoising is off versus an infeasible cost matrix in RTDetrLoss." + "reason": "Auxiliary-loss behavior and infeasible cost-matrix crash are different detection bugs." }, { - "left": "issue:27957", - "right": "issue:34689", + "left": "issue:44464", + "right": "issue:46424", "accept": false, - "reason": "Both involve model loading/attention support, but one is XLM-Roberta Flash Attention 2 support and the other is a Llama 3.2 Vision loading regression." + "reason": "Chunked generation inconsistencies and cache max-length handling are separate generation/cache issues." }, { - "left": "issue:44464", - "right": "issue:46424", + "left": "issue:27957", + "right": "issue:34689", "accept": false, - "reason": "Both mention generation/cache behavior, but one is compiled chunked generation inconsistency and the other is cache length configuration being ignored outside quantized cache." + "reason": "Flash Attention 2 support for XLMRoberta and Llama 3.2 Vision loading regression are not the same issue." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 zero-probability outputs are a model-specific inference issue, not the same as unexpected position-id keys across many models." + "reason": "Unexpected zero probabilities in SigLIP2 and unexpected key/position-id issues do not match." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to output embeddings being reinitialized after resizing token embeddings." + "reason": "Failing Flash Attention 2 tests and output embedding reinitialization after resizing are unrelated." }, { "left": "issue:29870", "right": "issue:32090", "accept": false, - "reason": "Bart generate-time decoder attention mask bug is unrelated to the Trainer `_gpu_broadcast_one` NoneType broadcast error." + "reason": "BART generate decoder_attention_mask assignment and Trainer broadcast error are different bugs." }, { "left": "issue:10000", "right": "issue:15354", "accept": false, - "reason": "Completely different eras and bugs: German DistilBertModel issue versus a torch.jit.script GeneratorExp export problem for ViT." + "reason": "German DistilBertModel issue and torch.jit.script GeneratorExp unsupported are unrelated." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "Both are dtype/vision-language bugs, but one is BF16-to-FP32 evaluation drift and the other is undefined video-token helper variables in VLM processors." - }, - { - "left": "issue:37428", - "right": "issue:42175", - "accept": false, - "reason": "Missing `_flash_supports_window_size` import and TensorFlow not included in `.[torch]` are unrelated packaging/API issues." + "reason": "Embedding dtype drift during evaluate and undefined video-token helper variables are different issues." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Both are chat/VL parsing problems, but one is tool-call assistant messages with no content and the other is Qwen2.5-VL video `vision_position_ids` behavior." - }, - { - "left": "issue:44448", - "right": "issue:44977", - "accept": false, - "reason": "Both are generation-output complaints, but they concern different models and different mechanisms (Pegasus regression vs Qwen3.5 flash-attention behavior)." + "reason": "Chat-template tool-call crash and qwen2.5-vl video position-id bug are separate code paths." }, { - "left": "issue:29870", - "right": "issue:33290", + "left": "issue:37428", + "right": "issue:42175", "accept": false, - "reason": "Bart decoder attention mask failure is unrelated to AdaFactor OOM in DeepSpeed." + "reason": "Missing _flash_supports_window_size import and missing TensorFlow in the torch extra are unrelated packaging/import problems." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is well grounded in the packet: the only clearly supported duplicate pair is issue:45310/issue:45478, which share the same title and appear to be the same Qwen3.5 MoE from_pretrained regression. The other soft-edge verdicts are conservative and avoid overclaiming sameness where the titles/body cues point to different bugs." + "accept": false, + "feedback": "The duplicate call on issue:45310 / issue:45478 is plausible, but the writeup overstates the evidence. The packet only shows identical titles, close dates, and similar body lengths; it does not justify \u201cnear-verbatim\u201d or \u201cmore complete.\u201d Please keep the canonical rationale more conservative, e.g. earlier report with matching title and same model/error family." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -10140,6 +10118,7 @@ "nodes": [ "issue:25251", "issue:29127", + "issue:29870", "issue:32090", "issue:33290", "issue:33357", @@ -10187,17 +10166,19 @@ "issue:46858" ], "soft_pairs": [ - "issue:44360|issue:44561", + "issue:44448|issue:44977", + "issue:29870|issue:33290", "issue:44855|issue:46399", + "issue:44360|issue:44561", "issue:43278|issue:43381", - "issue:45920|issue:46192", - "issue:32090|issue:33290", "issue:40444|issue:43526", + "issue:32090|issue:33290", + "issue:45920|issue:46192", "issue:45290|issue:46655", "issue:43927|issue:45698", "issue:44977|issue:45229", - "issue:46421|issue:46858", "issue:42490|issue:43873", + "issue:46421|issue:46858", "issue:43873|issue:45229", "issue:25251|issue:32090", "issue:44448|issue:44466", @@ -10208,217 +10189,229 @@ "issue:33357|issue:39692", "issue:43278|issue:43575", "issue:44561|issue:44779", - "issue:45588|issue:45823", "issue:44568|issue:45491", "issue:44361|issue:45198", + "issue:45588|issue:45823", "issue:41720|issue:45229", "issue:44186|issue:45198", "issue:45198|issue:45803", "issue:44977|issue:45563", "issue:43611|issue:46121" ], - "prepared_review_unit_hash": "b75a6a62b2dd9077e514a74aa3f794aa46e5af83d729b15ae4800fe67e760dc0", + "prepared_review_unit_hash": "471fb51b7fa2756b7b42416a8a8a41fe9ea45cee50c7a76786a91b1279051c50", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 36078, - "estimated_input_tokens": 9020, - "estimated_eval_tokens": 18296 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37315, + "estimated_input_tokens": 9329, + "estimated_eval_tokens": 18914 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a grab bag of mostly unrelated Transformers issues across tokenizers, generation, training, quantization, and model-specific regressions. None of the proposed soft pairs look like the same underlying bug or change, so all soft edges should be rejected.", + "summary": "These items are mostly unrelated standalone issues across different models, tokenizer paths, and runtime subsystems. The only broadly representative issue is the v5 AutoTokenizer regression affecting multiple model families.", "confidence": 0.93, - "canonical_issue_reason": "issue:43873 is the best anchor among these because it is open, has the strongest activity/inbound-reference signal, and describes a concrete but broad bug class (offloading + quantization) that is likely to attract related reports.", + "canonical_issue_reason": "issue:45920 has the broadest scope and strongest evidence of a shared underlying bug: a v5 regression affecting token IDs across multiple model families, with inbound references and clearer generality than the other issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the most suitable overall representative of the collection: it is active, externally referenced, and sufficiently substantive to serve as a triage target even though the rest of the set is largely unrelated.", + "best_issue_reason": "issue:45920 is the best representative issue because it covers a concrete, widely-impacting tokenizer regression rather than a model-specific edge case.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44360", - "right": "issue:44561", + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and different failure modes: Pegasus output drift vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:29870", + "right": "issue:33290", "accept": false, - "reason": "Different subsystems and failure modes: DSA indexing/ReLU discussion vs trust_remote_code loading breakage from a removed helper." + "reason": "Unrelated bugs: BART generate() unbound local error vs deepspeed/Adafactor OOM." }, { "left": "issue:44855", "right": "issue:46399", "accept": false, - "reason": "Both are import/runtime errors, but they affect different models and root causes (Python 3.13 torch.jit parsing vs missing registration decorator in Qwen3.5 kernels)." + "reason": "Both are import/runtime errors, but on different models and different code paths; not the same underlying bug." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "Completely different areas: DSA indexer discussion vs v5 trust_remote_code FX helper removal." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "One is a dtype regression between train/eval; the other is a hard restriction on checkpointing in eval mode. Not the same bug." + "reason": "Both touch eval mode, but one is a dtype regression and the other is a gradient-checkpointing restriction." }, { - "left": "issue:45920", - "right": "issue:46192", + "left": "issue:40444", + "right": "issue:43526", "accept": false, - "reason": "Tokenizer ID regression across several models vs return_tensors being ignored when text_kwargs is passed. Different API paths and symptoms." + "reason": "Different vision preprocessing bugs in different components/models." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "Trainer broadcast NoneType error and DeepSpeed Adafactor OOM are unrelated training-time failures." + "reason": "Both involve training infrastructure, but one is a NoneType broadcast error and the other is an optimizer memory issue." }, { - "left": "issue:40444", - "right": "issue:43526", + "left": "issue:45920", + "right": "issue:46192", "accept": false, - "reason": "Different processors and bugs: multi-image IterableDataset finetuning failure vs BeitImageProcessorFast label reduction behavior." + "reason": "Both are tokenizer-related, but one is a broad wrong-token-ID regression and the other is a return_tensors handling bug." }, { "left": "issue:45290", "right": "issue:46655", "accept": false, - "reason": "Chat template tool-call crash and SmolVLM video padding bug are separate preprocessing issues with different code paths." + "reason": "Different subsystems and symptoms: chat template/tool-call crash vs video processor padding calculation." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Both involve save/load, but one loses custom token IDs in config while the other loads the wrong custom module. Not the same underlying defect." + "reason": "Both involve save/load behavior, but one is config token IDs and the other is wrong custom module loading." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation bug for Qwen3.5 vs multi-GPU CUDA OOM for Gemma4; different models and failure mechanisms." - }, - { - "left": "issue:46421", - "right": "issue:46858", - "accept": false, - "reason": "Both mention cache/compile behavior, but one is lazy-init recompilation and the other is a device-tensor seq_length arithmetic bug on non-CUDA backends." + "reason": "Unrelated model-specific generation bug vs multi-GPU OOM." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Local-path trailing-slash import warning is unrelated to quantization/offloading behavior." + "reason": "Path parsing warning is unrelated to quantization/offloading behavior." + }, + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache/compile, but one is spurious recompilation and the other is a tensor-type arithmetic bug in prefill." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Quantization/offloading problems are not the same as a Gemma4 multi-GPU inference OOM, even if both can impact memory use." + "reason": "Different resource issues: quantization/offloading mismatch vs Gemma4 multi-GPU OOM." }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "Pipeline top_k nesting regression is unrelated to Trainer broadcast errors." + "reason": "Pipeline top_k nesting bug is unrelated to trainer GPU broadcast NoneType error." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "Different bugs: output mismatch for a summarization model vs serialization of tied lm_head weights." + "reason": "Both are v5-related, but one is generation output divergence and the other is serialization of lm_head.weight." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Eval/checkpointing restriction vs resume_from_checkpoint key mismatch; no shared root cause." + "reason": "Different training-state bugs: eval-mode checkpointing vs resume checkpoint key mismatch." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python version compatibility issue vs divergence between native and remote DeepSeekV3 implementations." + "reason": "One is a Python version compatibility regression, the other a model implementation divergence." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Tokenizer/model eos_token_id mismatch vs flash-attention generation failure for a different model family." + "reason": "Different model families and different issues: eos_token_id mismatch vs flash-attention generation failure." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "Error-message clarity issue in LayoutLMv3 vs TrainerState metric not updating. Different concerns." + "reason": "LayoutLMv3 error-message clarity is unrelated to TrainerState token counter updates." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS bus error on CLIP vs documentation example issues for SigLIP2; unrelated." + "reason": "Both are vision-model related, but one is a MacOS bus error and the other is a documentation-example mismatch/quantization failure." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "Embedding dtype change between train/eval is unrelated to tensor-parallel OOM when loading a large Qwen model." + "reason": "Different problems: eval dtype drift vs tensor-parallel load OOM." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "Removed helper breaking trust_remote_code models vs DeepSeek tokenizer regression. Both are v5-related but not the same bug." - }, - { - "left": "issue:45588", - "right": "issue:45823", - "accept": false, - "reason": "Flash-attention crash on s_aux=None and Gemma4 PLE device mismatch are distinct forward-path issues." + "reason": "Both are v5 regressions, but one breaks remote-code model loading and the other is tokenizer output correctness for DeepSeek." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Special-token insertion regression in mdeberta-v3-base vs NaN embeddings in Gemma3 sliding-window batching; different tokenizer/model paths." + "reason": "Tokenizer special-token insertion bug is unrelated to Gemma3 NaN embeddings on mixed-length batches." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "MLukeTokenizer AttributeError and Wav2Vec2 save_pretrained/tokenization failure are different model-specific bugs." + "reason": "Different tokenizer/processor bugs in different model families and code paths." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Both are forward/runtime model issues, but they affect different mechanisms and are not the same concrete bug." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Auto device mapping cudaErrorAssert and multi-GPU OOM are not the same concrete failure." + "reason": "Different hardware/resource failures: cudaErrorAssert with auto device mapping vs multi-GPU OOM." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "LayoutLMv2Tokenizer NER/padding crash vs Wav2Vec2 save_pretrained/tokenization failure; separate tokenizer issues." + "reason": "Different model/tokenizer failure modes; no evidence of a shared underlying bug." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Tokenization/save_pretrained failure vs broad exception swallowing in FuyuBatchFeature; different code and symptoms." + "reason": "Tokenization/save-pretrained failure is unrelated to a broad exception-handling bug in FuyuBatchFeature." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Generation failure with flash-attention vs a stale warning in paged generate; not the same bug." + "reason": "Generation crash vs a stale warning; not the same issue." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "Base_model_prefix loading breakage vs rope-validation helper rejecting a JSON-loaded list; unrelated loading/validation paths." + "reason": "Base-model-prefix loading regression is unrelated to rope-parameter validation on JSON-loaded lists." } ] }, @@ -10502,12 +10495,12 @@ "issue:44265|issue:45198", "issue:43575|issue:43701", "issue:29127|issue:33357", - "issue:43412|issue:45588", "issue:43065|issue:45412", "issue:43065|issue:43994", "issue:45491|issue:46693", - "issue:45372|issue:45588", + "issue:43412|issue:45588", "issue:42490|issue:43381", + "issue:45372|issue:45588", "issue:29127|issue:35141", "issue:45245|issue:45563", "issue:45459|issue:45712", @@ -10516,8 +10509,8 @@ "issue:43994|issue:45412", "issue:43575|issue:44387", "issue:43421|issue:45491", - "issue:45372|issue:45406", "issue:45406|issue:46701", + "issue:45372|issue:45406", "issue:43066|issue:44462", "issue:45412|issue:46369", "issue:44442|issue:45072", @@ -10525,240 +10518,247 @@ "issue:27957|issue:33290", "issue:43526|issue:45834", "issue:27957|issue:29127", - "issue:43526|issue:46693" + "issue:43526|issue:46693", + "issue:33290|issue:34689" ], - "prepared_review_unit_hash": "a9e27f8eb0fde8dbe67102706f7c27ae58d8e48cc791c6bf92e3438360275246", + "prepared_review_unit_hash": "2a36c6592ef7a953f048ef39f79c640d1feece4e32a5ee56e09d5c2e225bef27", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37835, - "estimated_input_tokens": 9459, - "estimated_eval_tokens": 19174 + "soft_pair_count": 35, + "serialized_chars": 38136, + "estimated_input_tokens": 9534, + "estimated_eval_tokens": 19324 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the issues span unrelated bugs across tokenizers, vision, generation, quantization, export, memory, and model-specific regressions. None of the soft pairs look like true duplicates or the same fixable code-path problem.", - "confidence": 0.94, - "canonical_issue_reason": "None of these issues look like the same underlying bug, so there is no true canonical issue for the cluster.", - "canonical_pr_reason": null, - "best_issue_reason": "Issue 46693 is the strongest anchor only by discussion activity and recency, but it still does not represent the rest of the heterogeneous set.", - "best_pr_reason": null, + "summary": "These items are mostly unrelated bugs across different models, tokenizers, processors, export paths, and memory behaviors. The soft-similarity pairs share broad subsystem words, but not the same concrete defect, so none should be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "No strong canonical issue: the set is heterogeneous and the suggested pairs do not indicate a single underlying bug family.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "issue:44462 is one of the clearest, most concrete reports here, with a specific reproducible loading bug and direct impact on AutoTokenizer behavior.", + "best_pr_reason": "No pull requests are present in this cluster.", "soft_edge_verdicts": [ { "left": "issue:29870", "right": "issue:34689", "accept": false, - "reason": "Different failures in different models and code paths: BART generate() mask handling vs Llama 3.2 vision model loading." + "reason": "Different failures in different areas: BART generate() decoder mask handling vs Llama 3.2 Vision model loading regression." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to image processor segmentation-map handling." + "reason": "Flash Attention 2 test failures and void segmentation-map processing are unrelated bugs." }, { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "Tokenizer special-token registration is unrelated to Qwen3.5 recurrent state reset behavior." + "reason": "Tokenizer special-token registration/empty decode is unrelated to Qwen recurrent state reset logic." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Unsupported feature-extractor arg for LasrFeatureExtractor is not the same bug as the Qwen video-input regression." + "reason": "Feature extractor center-arg regression and video-input regression affect different code paths and models." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "FineGrainedFP8Config loading failure is different from quantized model saving." + "reason": "Both involve quantization, but one is a FineGrainedFP8Config failure and the other is quantized saving; different concrete problems." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch is a model implementation bug, not the same as DeepSeekV3MoE divergence." + "reason": "Different model-implementation mismatches: Smollm3 RoPE layer dropping vs DeepseekV3MoE divergence from remote behavior." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer special-token registration is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + "reason": "BigBird tokenizer special-token registration bug is not the same as FastSpeech2ConformerTokenizer loading failure." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "torch.export failures with torch_compilable_check are not the same as Wav2Vec2 save_pretrained/tokenization breakage." + "reason": "torch.export with torch_compilable_check is a different failure mode from Wav2Vec2 save_pretrained/tokenization issues." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "Tensor-parallel OOM for Qwen2-57B is unrelated to resume_from_checkpoint key mismatch." + "reason": "OOM during tensor-parallel model load is unrelated to resume_from_checkpoint key mismatch." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "LayoutLMv3 error-message clarity and MacOS CLIP bus error are different issues with no shared code path." - }, - { - "left": "issue:43412", - "right": "issue:45588", - "accept": false, - "reason": "Gemma3n export issues are unrelated to flash_attention.py crashing on sink-less models." + "reason": "LayoutLMv3 error-message clarity and MacOS bus error on CLIP loading are unrelated." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory not being released on deletion." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory release/leak behavior." }, { "left": "issue:43065", "right": "issue:43994", "accept": false, - "reason": "Sam3PixelDecoder dummy layer and SigLIP2 nonsensical outputs are different model bugs." + "reason": "Sam3PixelDecoder dummy layer and SigLIP2 nonsensical outputs are different bugs in different models." }, { "left": "issue:45491", "right": "issue:46693", "accept": false, - "reason": "Both mention FA2 context behavior, but one is NaN embeddings in Gemma3 while the other is a performance regression; not the same concrete bug." + "reason": "Both mention long-context/FA2, but one is a Gemma3 NaN embedding bug and the other is a performance regression; not the same defect." }, { - "left": "issue:45372", + "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Gemma4 processor loading import failure is unrelated to flash_attention.py's s_aux=None crash." + "reason": "Gemma3n export guard/erfinv support and flash_attention s_aux=None AttributeError are distinct code-path failures." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Trailing-slash local model path warnings have nothing to do with gradient checkpointing in eval mode." + "reason": "Local-path import warning and gradient-checkpointing-in-eval-mode are unrelated." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 processor loading import error is unrelated to flash_attention sink-less model handling." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "LayoutLMv3 error messaging and token-embedding reinitialization after resize are unrelated." + "reason": "Error-message clarity in LayoutLMv3 is unrelated to post_init reinitializing resized embeddings." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "Category-cardinality RuntimeError and stale num_return_sequences warning are different generate/save behaviors." + "reason": "Category-cardinality RuntimeError and stale paged-generate warning are different issues." }, { "left": "issue:45459", "right": "issue:45712", "accept": false, - "reason": "Tokenizer protobuf error masking is unrelated to leftover dummy classes leaking into the package namespace." + "reason": "Tokenizer protobuf exception masking and leftover dummy PT classes are unrelated." }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "DeepSeekV3MoE implementation divergence is unrelated to D-FINE auxiliary-loss behavior." + "reason": "DeepseekV3MoE divergence and D-FINE auxiliary-loss behavior are different model bugs." }, { "left": "issue:43526", "right": "issue:47217", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is different from VL placeholder-handling failures." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to VL image-placeholder handling." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 wrong outputs and RT-DETR memory cleanup are unrelated problems." + "reason": "SigLIP2 wrong outputs and RT-DETR memory release are unrelated defects." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Both mention OOM, but one is tensor-parallel model loading and the other is int4 CUDA reserved-memory regression; different causes and fixes." + "reason": "Both can OOM, but tensor-parallel load OOM and int4 quantization reserved-memory regression have different root causes." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Tokenizer post-processor updates at runtime are unrelated to Gemma3 NaN embeddings in sliding-window attention." + "reason": "Runtime special-token post-processor update bug is unrelated to Gemma3 mixed-length NaN embeddings." }, { - "left": "issue:45372", - "right": "issue:45406", + "left": "issue:45406", + "right": "issue:46701", "accept": false, - "reason": "Gemma4 processor import failure and Gemma4Processor missing _tokenizer in serve are distinct failures." + "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to DiffusionGemma image-text-to-text pipeline failure." }, { - "left": "issue:45406", - "right": "issue:46701", + "left": "issue:45372", + "right": "issue:45406", "accept": false, - "reason": "Gemma4 processor serve crash is unrelated to DiffusionGemmaGenerationOutput pipeline failure." + "reason": "Gemma 4 processor loading import error and Gemma4Processor missing _tokenizer are different failures." }, { "left": "issue:43066", "right": "issue:44462", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring tokenizer.json." + "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring tokenizer.json from the repository." }, { "left": "issue:45412", "right": "issue:46369", "accept": false, - "reason": "Garbage-collection memory retention and a SAM3-litetext memory overflow are different memory bugs." + "reason": "RT-DETR memory leak and Sam3-litetext memory overflow are different memory issues in different models." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "FastSpeech2Conformer tokenizer loading is unrelated to dtype mismatches in bfloat16 inference." + "reason": "FastSpeech2ConformerTokenizer loading and bfloat16 dtype mismatches are unrelated." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "Missing ReasoningEffort import for Gemma4 loading is unrelated to protobuf decode-error masking." + "reason": "Gemma 4 import failure from mistral_common and protobuf decode-error masking affect different tokenizer/processor paths." }, { "left": "issue:27957", "right": "issue:33290", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Adafactor/deepspeed OOM are different issues." + "reason": "XLMRoberta Flash Attention 2 support and Adafactor+DeepSpeed OOM are unrelated." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "BeitImageProcessorFast label reduction is unrelated to Kosmos2.5 long OCR index errors." + "reason": "BeitImageProcessorFast label reduction and Kosmos2.5 long OCR index error are unrelated." }, { "left": "issue:27957", "right": "issue:29127", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta and LayoutLMv3 error messaging have no shared bug path." + "reason": "Flash Attention 2 support for XLMRoberta and LayoutLMv3 error messaging have no common underlying bug." }, { "left": "issue:43526", "right": "issue:46693", "accept": false, - "reason": "Label reduction in image processing is unrelated to FA2 long-context generation performance regression." + "reason": "Label reduction bug and FA2 long-context performance regression are unrelated." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "DeepSpeed Adafactor OOM and Llama 3.2 Vision loading breakage are different issues." } ] }, @@ -10785,7 +10785,6 @@ "issue:27957", "issue:29870", "issue:33290", - "issue:34689", "issue:35141", "issue:42175", "issue:42490", @@ -10833,20 +10832,19 @@ ], "soft_pairs": [ "issue:43824|issue:44279", - "issue:33290|issue:34689", "issue:33290|issue:35141", "issue:44857|issue:45588", - "issue:43334|issue:43531", "issue:42915|issue:43931", "issue:43994|issue:46369", + "issue:43334|issue:43531", "issue:45250|issue:46640", "issue:43381|issue:43873", "issue:45588|issue:46821", - "issue:43646|issue:44077", "issue:46121|issue:46399", "issue:44514|issue:45381", - "issue:43761|issue:45859", + "issue:43646|issue:44077", "issue:44977|issue:45245", + "issue:43761|issue:45859", "issue:44743|issue:44811", "issue:44442|issue:44898", "issue:43262|issue:44206", @@ -10854,202 +10852,196 @@ "issue:43720|issue:44442", "issue:27957|issue:29870", "issue:42175|issue:46234", + "issue:44077|issue:44479", "issue:43316|issue:43525", "issue:45003|issue:45440", - "issue:44077|issue:44479", "issue:43493|issue:43827", "issue:43873|issue:44871" ], - "prepared_review_unit_hash": "f444ba2bd54e7643b4fea5d4d329ae445ad7fd0b57e56d071c272c06fdbbd876", + "prepared_review_unit_hash": "016d4ef93c18ed0e53b698668ddf457eeb8b7b110b0c1b2dc075698bf883e5f2", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35688, - "estimated_input_tokens": 8922, - "estimated_eval_tokens": 18100 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34814, + "estimated_input_tokens": 8704, + "estimated_eval_tokens": 17664 }, "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 soft links are largely superficial keyword/model-family overlaps rather than the same underlying defect or change, so I do not see any safe duplicate merges.", - "confidence": 0.86, - "canonical_issue_reason": "No single canonical issue stands out: the items span distinct code paths (loading, generation, quantization/offloading, audio, vision, docs, packaging), so there is no clear duplicate center.", + "summary": "This cluster is largely heterogeneous: the soft pairs look like distinct bugs or regressions in different code paths, so I would reject every soft edge. There are no pull requests in the cluster. If one issue has to serve as a representative, 43873 is the most central/active item, but it is not a true duplicate umbrella for the rest.", + "confidence": 0.41, + "canonical_issue_reason": "issue:43873 is the best representative only because it is active, has inbound references, and describes a broad runtime behavior regression; however, the cluster does not form one coherent duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the most representative/open report in the set: it has the most discussion/inbound references and a broad active theme around quantization/offloading.", + "best_issue_reason": "issue:43873 is the most suitable issue to anchor triage because it is open, discussion-heavy, and broader than the other items, which are mostly isolated model/config regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "Different problems: Qwen2.5-VL import failure vs a generic dependency/install issue." - }, - { - "left": "issue:33290", - "right": "issue:34689", - "accept": false, - "reason": "Adafactor/DeepSpeed OOM is unrelated to a Llama 3.2 Vision model-loading regression." + "reason": "One is a Qwen2.5-VL import error; the other is a vague dependency issue. No shared failing code path." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "Optimizer memory usage and embedding reinitialization in post_init are different bugs." + "reason": "AdaFactor/DeepSpeed OOM and embedding reinitialization in post_init are unrelated bugs." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "AMP/CUDA loss crash vs flash-attention sinkless AttributeError; different code paths." - }, - { - "left": "issue:43334", - "right": "issue:43531", - "accept": false, - "reason": "Missing pad_token_id on Qwen3-VL config is not the same as sliding_window handling in Qwen3-MoE." + "reason": "Different failures in different subsystems: LwDETR AMP/CUDA loss vs flash-attention sink-less model crash." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "FP8 config failure in Qwen3Moe vs weight-shape mismatch for Qwen3-VL-30B are distinct issues." + "reason": "Qwen3MoE FP8 config failure and Qwen3-VL weight-shape mismatch are distinct model-loading problems." }, { "left": "issue:43994", "right": "issue:46369", "accept": false, - "reason": "SigLIP2 wrong outputs and SAM3 memory overflow are unrelated failures." + "reason": "SigLIP2 incorrect outputs and SAM3 memory overflow are different models and different symptoms." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id in Qwen3-VL config is not the same as a sliding_window issue in Qwen3-MoE." }, { "left": "issue:45250", "right": "issue:46640", "accept": false, - "reason": "Both mention Flash Attention 2, but one is a generic note and the other is a specific crash report; not the same bug." + "reason": "Both mention Flash Attention 2, but the titles are too generic to show the same concrete bug." }, { "left": "issue:43381", "right": "issue:43873", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to quantization/offloading behavior." + "reason": "Gradient checkpointing in eval mode and quantization/offloading behavior are separate mechanisms." }, { "left": "issue:45588", "right": "issue:46821", "accept": false, - "reason": "Flash-attention AttributeError and TimesFM window_size AttributeError affect different components and code paths." - }, - { - "left": "issue:43646", - "right": "issue:44077", - "accept": false, - "reason": "Both touch initialization/post_init conceptually, but one is a v5 custom model init break and the other is a PatchTSMixer-specific policy issue." + "reason": "A flash-attention AttributeError and a TimesFM window_size AttributeError are unrelated failures." }, { "left": "issue:46121", "right": "issue:46399", "accept": false, - "reason": "Rope validation TypeError and missing decorator for rotary kernel registration are different defects." + "reason": "Rope-validation TypeError and Qwen3.5 apply_rotary_pos_emb registration failure are different code paths." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Batched apply_chat_template padding failure is separate from the qwen2.5-vl video vision_position_ids regression." + "reason": "Both are Qwen2.5-VL regressions, but one is batched chat-template padding and the other is video vision_position_ids." }, { - "left": "issue:43761", - "right": "issue:45859", + "left": "issue:43646", + "right": "issue:44077", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and Qwen rotary embedding CPU-offload incompatibility are unrelated." + "reason": "A broad v5 custom-init break and a patchtsmixer post_init policy issue are not the same concrete change." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Flash-attention generation issues do not match the category-count runtime error." + "reason": "Qwen3.5 flash-attention generation failure and category-count overflow are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are unrelated." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Recurrent-state reset with cache is unrelated to Whisper batch_decode skipping special tokens." + "reason": "Qwen3.5 recurrent-state reset and Whisper batch_decode skip_special_tokens are different features." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "FastSpeech2Conformer tokenizer loading failure is not the same as Perceiver non-default-resolution failure." + "reason": "Tokenizer loading failure and Perceiver non-default resolution failure are unrelated." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Audio chat_template sampling-rate defaulting and unsupported center arg in LasrFeatureExtractor are different audio bugs." + "reason": "Audio chat-template sampling-rate default and LaserFeatureExtractor center-arg crash are different audio bugs." }, { "left": "issue:42490", "right": "issue:44743", "accept": false, - "reason": "Trailing-slash local-model import warning is unrelated to recurrent-state reset behavior." + "reason": "Local-path import warning and Qwen3.5 recurrent-state reset do not share a code path." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "BitNet packed-weight loading and tokenizer loading are different subsystems." + "reason": "BitNet packed-weight loading and tokenizer loading are unrelated." }, { "left": "issue:27957", "right": "issue:29870", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Bart generate() decoder_attention_mask initialization are unrelated." + "reason": "XLMRoberta Flash Attention 2 support and BART generate() decoder_attention_mask error are distinct issues." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Packaging/backend omission is not the same as a missing import_utils helper." + "reason": "Torch-only install backend packaging and missing import_utils symbol are not the same bug." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "PatchTSMixer post_init policy and Qwen multi-VL video regression are unrelated." }, { "left": "issue:43316", "right": "issue:43525", "accept": false, - "reason": "Different model config API discrepancies; not the same underlying bug." + "reason": "Gemma3TextConfig API discrepancy and Llama4Config missing pad_token_id are separate model config mismatches." }, { "left": "issue:45003", "right": "issue:45440", "accept": false, - "reason": "Unsafe sys.modules access and DeepseekV3MoE divergence are unrelated." - }, - { - "left": "issue:44077", - "right": "issue:44479", - "accept": false, - "reason": "PatchTSMixer post_init policy is unrelated to Qwen VL video-input regression." + "reason": "sys.modules access safety and DeepSeekV3MoE divergence are different problems." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation mismatch and docs still referencing removed pipeline APIs are unrelated." + "reason": "SigLIP2 implementation mismatch and outdated docs referencing pipeline() are unrelated." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading behavior and Gemma eos_token_id inconsistency are different problems." + "reason": "Quantization/offloading behavior and Gemma-3 eos_token_id configuration are different issues." } ] }, @@ -11127,15 +11119,15 @@ "issue:43643|issue:44913", "issue:45325|issue:46697", "issue:43316|issue:45759", - "issue:44987|issue:46178", "issue:43749|issue:43931", + "issue:44987|issue:46178", "issue:44190|issue:44496", - "issue:43701|issue:46506", "issue:43931|issue:45216", + "issue:43701|issue:46506", "issue:42617|issue:43575", "issue:44442|issue:45803", - "issue:43232|issue:43388", "issue:43550|issue:43701", + "issue:43232|issue:43388", "issue:43575|issue:43742", "issue:43761|issue:44079", "issue:27957|issue:32090", @@ -11150,20 +11142,20 @@ "issue:44521|issue:47217", "issue:44871|issue:45584" ], - "prepared_review_unit_hash": "6a59f4ab56a73a48a5c5c2722cc13cb34be03385ea248365f9ac16f5a08c6bfd", + "prepared_review_unit_hash": "44d4d6a34413369cb5529a4dfa3d30349466cc5748d969bff6f2217475d92dc2", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35291, + "serialized_chars": 35292, "estimated_input_tokens": 8823, "estimated_eval_tokens": 17902 }, @@ -11172,174 +11164,174 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is a grab bag of mostly unrelated Transformers issues; none of the soft pairs look like the same underlying bug or change, so no soft edge should be accepted. There are no PRs in this cluster; the strongest issue anchor is the high-activity FSDP loading regression.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43749 has the strongest triage signal in this set: high discussion/inbound references, clear regression wording, and a concrete subsystem failure (FSDP CPU RAM efficient loading).", + "summary": "The items span many unrelated bugs and regressions, with no soft pair looking like the same underlying change strongly enough to merge. I\u2019d keep the cluster anchored on the most actionable, high-signal issue rather than treating any pair as duplicates.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43749 stands out as the clearest high-signal regression: it has the most discussion and inbound references, and the title points to a concrete broken feature rather than a vague environment or usage question.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43749 is the best representative issue because it is high-impact, well-discussed, and specific enough to serve as an anchor for related follow-ups; the other issues are narrower or more isolated.", + "best_issue_reason": "issue:44521 is the best active triage anchor because it is open, has multiple inbound references, and describes a specific reproducible multimodal API failure.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44062", "right": "issue:45907", "accept": false, - "reason": "Different code paths and failure classes: tokenizer AddedToken TypeError vs Hub list_repo_files not forwarding revision." + "reason": "Different code paths: tokenizer AddedToken keyword collision vs list_repo_files not forwarding revision; unrelated bugs." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Python 3.13 torch.jit parsing/IndentationError vs DeepSpeed ZeRO3 deque underflow; unrelated subsystems." + "reason": "Python 3.13 torch.jit parsing/import error vs DeepSpeed ZeRO3 deque underflow; no shared underlying defect." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "Both are config-related, but one is missing fields from trust_remote_code loading and the other is rotary_pct not persisting on reload." + "reason": "One is missing fields from AutoConfig with trust_remote_code, the other is GPTNeoX config reload losing rotary_pct; different config bugs." }, { "left": "issue:45325", "right": "issue:46697", "accept": false, - "reason": "Different models and bugs: Qwen2.5-VL rope/position-id scaling vs Swin2SR image over-padding." + "reason": "Qwen2.5-VL rope/temporal position scaling issue vs Swin2SR over-padding in image processor; different preprocessing logic." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Gemma3TextConfig API discrepancy is not the same as AutoModelForCausalLM not unwrapping text_config for composite multimodal configs." + "reason": "Gemma3TextConfig API discrepancy vs Qwen multimodal from_config text_config unwrapping; different model families and failure modes." }, { - "left": "issue:44987", - "right": "issue:46178", + "left": "issue:43749", + "right": "issue:43931", "accept": false, - "reason": "Loading physical-intelligence/fast failure is unrelated to BACKENDS_MAPPING rejecting tensorflow_text." + "reason": "FSDP CPU RAM efficient loading regression is unrelated to a model-specific weight shape mismatch on load." }, { - "left": "issue:43749", - "right": "issue:43931", + "left": "issue:44987", + "right": "issue:46178", "accept": false, - "reason": "Both mention loading, but one is FSDP CPU RAM loading broken and the other is a Qwen3-VL weight-shape mismatch on a specific model." + "reason": "Loading a specific model on transformers>=5.1.0 vs tensorflow_text backend mapping error; no common bug." }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to AutoTokenizer failing to load a custom tokenizer." + "reason": "Dataset loading in an example script vs AutoTokenizer rejecting an unrecognized model config; unrelated entry points." }, { - "left": "issue:43701", - "right": "issue:46506", + "left": "issue:43931", + "right": "issue:45216", "accept": false, - "reason": "resume_from_checkpoint key mismatch is a checkpoint/state issue, while PPDocLayoutV3 CPU/CUDA differences are numerical consistency problems." + "reason": "Model weight shape mismatch on load vs corrupted save_pretrained checkpoint output; distinct load/save bugs." }, { - "left": "issue:43931", - "right": "issue:45216", + "left": "issue:43701", + "right": "issue:46506", "accept": false, - "reason": "One is a load-time tensor shape mismatch; the other is a save_pretrained regression producing incorrect checkpoints." + "reason": "resume_from_checkpoint key mismatch vs CPU/CUDA output divergence in PPDocLayoutV3; unrelated." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py execution failure is not the same as tp loading OOM for Qwen2-57B." + "reason": "Failure to run a parallel script vs OOM when loading a large model with tensor parallelism; same broad area, different problem." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer loading failure for FastSpeech2Conformer is unrelated to FuyuBatchFeature swallowing exceptions." + "reason": "Tokenizer loading failure vs bare except swallowing interrupts in Fuyu tensor conversion; unrelated." }, { - "left": "issue:43232", - "right": "issue:43388", + "left": "issue:43550", + "right": "issue:43701", "accept": false, - "reason": "Generation kwargs/update bug and gather_for_metrics label truncation are different training/inference paths." + "reason": "torch.compile SDPA issue in Bamba vs checkpoint resume key mismatch; different subsystems." }, { - "left": "issue:43550", - "right": "issue:43701", + "left": "issue:43232", + "right": "issue:43388", "accept": false, - "reason": "torch.compile SDPA divergence in Bamba is unrelated to resume_from_checkpoint key mismatch." + "reason": "Generation kwargs sync-gpus issue vs gather_for_metrics dropping tuple labels; different distributed-training bugs." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "Qwen2-57B tensor-parallel OOM is a different failure from loading MobileLLM-125M with a key error." + "reason": "OOM loading a large Qwen model vs key error loading MobileLLM-125M; not the same underlying issue." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and ModelOutput key assignment bug are distinct output-handling issues." + "reason": "CLIPVisionModel hidden_states regression vs ModelOutput key assignment behavior when previous key is None; separate output-structure bugs." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support issue is unrelated to Trainer _gpu_broadcast_one receiving NoneType." + "reason": "Flash Attention 2 support for XLMRoberta vs Trainer broadcast TypeError on None; unrelated." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "One is a model forward output regression; the other is attention-kernel divergence under torch.compile." + "reason": "Missing hidden_states output vs attention kernel divergence under torch.compile; different code paths." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch is unrelated to resume_from_checkpoint key mismatch." + "reason": "CPU offload device mismatch in bark-small vs resume_from_checkpoint key mismatch; unrelated." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "pipeline top_k nesting behavior and missing PyTorch installation are different problems." + "reason": "pipeline top_k output shape regression vs missing PyTorch dependency error; unrelated." }, { "left": "issue:46829", "right": "issue:46932", "accept": false, - "reason": "VideoPrism hidden_states mapping bug and Gemma4 dtype casting for quantized embedders are separate model-specific regressions." + "reason": "VideoPrism hidden_states misassignment vs Gemma4 quantized embedder dtype casting fix; distinct model internals." }, { "left": "issue:42907", "right": "issue:45698", "accept": false, - "reason": "Dequantized model saving failure and wrong custom module loading after save_pretrained affect different persistence paths." + "reason": "Saving dequantized models correctly vs loading the wrong custom module after save_pretrained; both serialization-related but not the same bug." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "A model_max_length question for DeBERTa-V3 is unrelated to the missing transformers.onnx import error." + "reason": "model_max_length question for DeBERTa-V3 vs missing transformers.onnx module import; unrelated." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "Processor/tokenizer regression and Gemma-3 eos_token_id inconsistency are different config/API mismatches." + "reason": "processor.tokenizer/image-to-tokenizer regression vs Gemma-3 eos_token_id config mismatch; separate regression and config issues." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Custom processor.tokenizer access regression is unrelated to a pipeline error-message wording issue." + "reason": "Custom multimodal processor regression vs pipeline error message about translation support in v5; different user-facing failures." }, { "left": "issue:44521", "right": "issue:47217", "accept": false, - "reason": "Both are multimodal, but one is all-zero assistant masks from apply_chat_template and the other is failure when the image placeholder appears in user text." + "reason": "Both involve multimodal prompting, but one is all-zero assistant masks and the other is failure when an image placeholder appears in user text; not clearly the same defect." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Gemma-3 eos_token_id mismatch and Whisper empty-transcription generation failure are unrelated model-specific issues." + "reason": "Gemma-3 eos_token_id inconsistency vs Whisper empty transcription after special-token alignment; unrelated special-token bugs." } ] }, @@ -11414,10 +11406,10 @@ "soft_pairs": [ "issue:46234|issue:46291", "issue:43316|issue:43716", - "issue:44077|issue:44534", "issue:43316|issue:46821", "issue:44464|issue:46421", "issue:43646|issue:44913", + "issue:44077|issue:44534", "issue:43452|issue:44496", "issue:44871|issue:46050", "issue:45803|issue:46365", @@ -11425,34 +11417,34 @@ "issue:43404|issue:43441", "issue:44242|issue:46894", "issue:43065|issue:44617", - "issue:43572|issue:46178", "issue:45632|issue:46424", + "issue:43572|issue:46178", "issue:43540|issue:44514", "issue:44610|issue:46369", "issue:45381|issue:47217", "issue:45542|issue:46234", "issue:44792|issue:45600", + "issue:43504|issue:45698", "issue:45803|issue:46170", "issue:44792|issue:44977", - "issue:43504|issue:45698", "issue:43352|issue:46640", "issue:45127|issue:45200", "issue:43525|issue:43531" ], - "prepared_review_unit_hash": "9470799050718542dea7d147a452da404848f99bc7778c64280e61aef1102110", + "prepared_review_unit_hash": "984f8bae52e47fe474a4bef743abbcf8453bdf95ec2cf5f379d20f669cfeb9e5", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 34754, + "serialized_chars": 34753, "estimated_input_tokens": 8689, "estimated_eval_tokens": 17634 }, @@ -11461,168 +11453,168 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab bag of unrelated closed issues across model configs, import/runtime environment failures, vision-language processors, cache behavior, and a few FlashAttention regressions. None of the soft pairs look like the same underlying bug/change strongly enough to merge.", - "confidence": 0.9, - "canonical_issue_reason": "issue:45698 is the strongest representative by discussion activity and inbound references, and it describes a concrete, user-facing save/load bug with broad impact.", + "summary": "The cluster is mostly a loose set of unrelated Transformers issue reports with superficial keyword overlap (FlashAttention, cache/save/load, multimodal/VL, v5 regressions). I do not see any strong duplicate pairs; all soft edges should be rejected.", + "confidence": 0.44, + "canonical_issue_reason": "issue:43646 is the broadest umbrella-style Transformers v5 regression in the set, so it is the closest single representative, though the cluster is fairly heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45698 is the most suitable issue to stand in for the set because it has the clearest reproduction, the highest observed discussion/references, and a direct model-loading failure mode.", + "best_issue_reason": "issue:45698 is the most concrete and well-scoped core loading bug, with strong discussion/inbound references, making it the best standalone issue to anchor triage if one item must be chosen.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46234", "right": "issue:46291", "accept": false, - "reason": "Different failures: missing import-utils availability helper vs pipelines import-time kernel-version requirement." + "reason": "Both involve optional dependencies, but one is a missing utility symbol and the other is a pipelines import failure tied to kernels version/revision requirements; different code paths." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Unrelated model issues: config API discrepancy vs image-preprocessor/model dtype mismatch." - }, - { - "left": "issue:44077", - "right": "issue:44534", - "accept": false, - "reason": "Both are v5 regressions, but they describe different bugs and different code paths." + "reason": "Gemma3TextConfig API mismatch and Mistral-3 image preprocessor dtype mismatch are unrelated bugs in different subsystems." }, { "left": "issue:43316", "right": "issue:46821", "accept": false, - "reason": "No shared bug: Gemma config parity issue vs TimesFM window_size attribute error." + "reason": "Config API discrepancy for Gemma3TextConfig is not the same as TimesFM window_size raising AttributeError." }, { "left": "issue:44464", "right": "issue:46421", "accept": false, - "reason": "Both mention chunking/compilation, but one is output inconsistency and the other is spurious recompilation from cache init." + "reason": "Both mention chunked/compiled execution, but one is inconsistent generation outputs and the other is cache lazy init causing recompilation; not the same bug." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "General custom-model init breakage is not the same as GPTNeoX rotary_pct reloading/reset behavior." + "reason": "A broad v5 custom initialization regression is different from GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init validation for patchtsmixer and non-persistent buffer junk-fill are separate v5 issues with different failure modes." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "Different root causes: gguf_file/from_pretrained handling vs unrecognized model config/model_type." + "reason": "gguf_file loading breakage and unrecognized-model/model_type errors are both loading problems but not the same underlying code-path." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "Gemma eos_token_id inconsistency and sinq quantized save/load breakage are distinct bugs." + "reason": "Gemma-3 eos_token_id inconsistency and sinq quantized save/load breakage concern different model metadata and serialization flows." }, { "left": "issue:45803", "right": "issue:46365", "accept": false, - "reason": "Bare except swallowing errors is unrelated to Trainer import failure caused by datasets." + "reason": "Bare except in a tensor conversion helper and Trainer import failure due to datasets are unrelated." }, { "left": "issue:44162", "right": "issue:44792", "accept": false, - "reason": "ESM2 breakage and Janus image-generation test failure are unrelated model-specific problems." + "reason": "ESM2 breakage and a janus image-generation test failure are different models and different failures." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Different Mistral-family issues: tied weights bug vs FlashAttention support failure." + "reason": "Weight tying in Mistral3ForConditionalGeneration is a different problem from Ministral-3 FlashAttention compatibility." }, { "left": "issue:44242", "right": "issue:46894", "accept": false, - "reason": "MoE load-balancing loss bug is not the same as CohereASR double-shift training-loss issue." + "reason": "Load-balancing loss omission when output_router_logits=False is not the same as the CohereASR double-shift training-loss bug." }, { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "Sam3PixelDecoder dummy conv and Sam3Video OOM are different symptoms and likely different fixes." + "reason": "Dummy Conv2d in Sam3PixelDecoder and Sam3Video CUDA OOM are not the same issue; one is architectural, the other resource usage." }, { - "left": "issue:43572", - "right": "issue:46178", + "left": "issue:45632", + "right": "issue:46424", "accept": false, - "reason": "Missing pad_token_idx in a config is unrelated to a backend mapping error for tensorflow_text." + "reason": "Trust-remote-code cache path collisions and max_cache_length only applying to Quantized Cache are separate cache/config bugs." }, { - "left": "issue:45632", - "right": "issue:46424", + "left": "issue:43572", + "right": "issue:46178", "accept": false, - "reason": "Both involve cache, but one is path collision in trust_remote_code and the other is max_cache_length semantics." + "reason": "StableLmConfig missing pad_token_idx and a backend mapping error for tensorflow_text are unrelated config vs import/backend issues." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Different multimodal bugs: video input processing vs batched chat template padding behavior." + "reason": "Qwen3OmniMoe video-input ValueError and Qwen2_5_VLProcessor batched chat-template crash have different entrypoints and failure conditions." }, { "left": "issue:44610", "right": "issue:46369", "accept": false, - "reason": "Processor size mismatch and Sam3-LiteText memory overflow are not the same underlying issue." + "reason": "OmDet-Turbo input-size mismatch and Sam3-litetext memory overflow are different symptoms and different model pipelines." }, { "left": "issue:45381", "right": "issue:47217", "accept": false, - "reason": "Both are VL issues, but one is video position IDs and the other is image-placeholder handling in user text." + "reason": "Wrong vision_position_ids for Qwen2.5-VL video inputs is not the same as image-placeholder handling in user text for VL models." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "TensorFlow backend probing issue is unrelated to the falcon-mamba availability helper missing from import_utils." + "reason": "TensorBoard without TensorFlow causing backend errors and a missing import_utils availability helper are different dependency issues." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "Janus test failure and removed Sam3LiteText auto-mapping references are unrelated CI/model issues." + "reason": "A janus generation test failure and auto_mappings still referencing removed Sam3LiteText configs are separate CI breakages." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy-field loading failure for BEiT semantic segmentation and wrong custom-module restoration after save_pretrained are distinct load-time bugs." }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Exception swallowing in tensor conversion is unrelated to Windows/ROCm import problems from torch.distributed dependency." + "reason": "Swallowing KeyboardInterrupt in a tensor conversion helper and a Windows/ROCm torch.distributed import problem are unrelated." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus image-generation test failure is not the same as Qwen3.5 FlashAttention generation issues." - }, - { - "left": "issue:43504", - "right": "issue:45698", - "accept": false, - "reason": "Legacy preset field loading and wrong custom-module resolution after save/load are different bugs." + "reason": "Janus test failure and Qwen3.5 FlashAttention generation problems are different models and different code paths." }, { "left": "issue:43352", "right": "issue:46640", "accept": false, - "reason": "Nemotron FlashAttention 2 unsupported and a generic Flash Attention 2 issue are too vague to treat as the same bug." + "reason": "NemotronHForCausalLM not supporting FlashAttention 2 yet is a specific model capability issue; the other is a generic flash-attention report without evidence of the same root cause." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "LoRA tied-embeddings collapse and Gemma 4 mm_token_type_ids defaulting are unrelated." + "reason": "LoRA merge collapse with extended vocabulary and Gemma 4 mm_token_type_ids defaults are unrelated training/config bugs." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Missing pad_token_id in Llama4Config and Qwen3-MoE sliding_window behavior are distinct config/model bugs." + "reason": "Missing pad_token_id in Llama4Config and Qwen3-MoE sliding_window issues are separate config/runtime problems." } ] }, @@ -11697,47 +11689,47 @@ ], "soft_pairs": [ "issue:43650|issue:46202", - "issue:44534|issue:44913", "issue:44479|issue:47217", - "issue:35141|issue:42490", + "issue:44534|issue:44913", "issue:41950|issue:43475", "issue:43673|issue:44743", + "issue:35141|issue:42490", "issue:43388|issue:43688", "issue:44190|issue:44964", "issue:46693|issue:47217", "issue:44855|issue:45588", "issue:29870|issue:35141", - "issue:45072|issue:46932", - "issue:44869|issue:46832", "issue:45632|issue:46617", + "issue:44869|issue:46832", + "issue:45072|issue:46932", "issue:45245|issue:46858", - "issue:44734|issue:45406", "issue:46082|issue:46697", + "issue:44734|issue:45406", "issue:43827|issue:43994", - "issue:44617|issue:45584", "issue:43644|issue:44913", + "issue:44617|issue:45584", "issue:25251|issue:27957", "issue:43335|issue:44855", "issue:43931|issue:44060", - "issue:44242|issue:46897", "issue:44077|issue:44877", + "issue:44242|issue:46897", "issue:43673|issue:46421", "issue:45563|issue:45584" ], - "prepared_review_unit_hash": "668253bb65fcf9c530214f27418e9c04fb46b4fba6b72dcfa2f3b897283c61e4", + "prepared_review_unit_hash": "3fc661e55383163d92f56ffc4a51bd854b8e5b2c9c15c537f43858f8d9c19f66", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36003, + "serialized_chars": 36001, "estimated_input_tokens": 9001, "estimated_eval_tokens": 18258 }, @@ -11746,174 +11738,174 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft edge is convincing enough to merge: the candidate pairs mostly share only broad keywords like cache, generation, pipeline, or vision, but they describe different bugs, models, or failure modes. This looks like a noisy mixed cluster rather than a duplicate set.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the items do not form a coherent duplicate group, so picking one representative would be arbitrary.", + "summary": "The proposed soft edges are mostly false positives: the paired issues touch different models, pipelines, cache behaviors, config bugs, or docs, and do not describe the same underlying bug/change. I do not see a clear canonical issue or PR for this set because the items are heterogeneous and none of the candidate pairs are strong duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a strong global representative because the cluster spans unrelated reports across many subsystems.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43650", "right": "issue:46202", "accept": false, - "reason": "A vague placeholder issue versus a Gemma pipeline roundtrip bug; unrelated." - }, - { - "left": "issue:44534", - "right": "issue:44913", - "accept": false, - "reason": "Non-persistent buffer junk vs GPTNeoX rotary_pct reload behavior; different code paths." + "reason": "Unrelated topics: one is a vague data request, the other is a Gemma pipeline roundtrip tokenizer edge case. No shared code path or concrete duplicate bug." }, { "left": "issue:44479", "right": "issue:47217", "accept": false, - "reason": "Both are VL-related, but one is a video-input regression and the other is an image-placeholder-in-user-text failure; not the same bug." + "reason": "Different regressions: v5.3.0 video-input model-family issue versus VL placeholder handling in user text. Similar domain, but not the same bug." }, { - "left": "issue:35141", - "right": "issue:42490", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "Embedding reinitialization on resize vs local-path import warning; unrelated." + "reason": "Both mention v5/config behavior, but one is about non-persistent buffers being initialized with junk and the other about GPTNeoXConfig rotary_pct not persisting on reload. Different failure modes and code paths." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Pipeline processor lookup bug vs missing fpn_position_embeddings on a Sam3 output object; unrelated." + "reason": "Video-classification pipeline processor lookup is unrelated to SAM3 vision encoder missing fpn_position_embeddings. Different subsystems and fixes." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both mention cache, but one is missing GenerationMixin cache in chunked prefill and the other resets recurrent states for seq_len>1; different failures." + "reason": "Chunked-prefill cache absence in GenerationMixin is not the same as recurrent states resetting in modular_qwen3_5.py. Different mechanisms and models." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "One is token embedding reinitialization after resize; the other is trust_remote_code path normalization with trailing slash. No overlap beyond generic loading behavior." }, { "left": "issue:43388", "right": "issue:43688", "accept": false, - "reason": "Metrics label-dropping bug vs auxiliary-loss normalization; unrelated." + "reason": "Both concern training/metrics/losses, but one is gather_for_metrics truncating tuple labels and the other is auxiliary loss normalization in MoE models. Distinct bugs." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Dataset loading in an example script vs multimodal model loading failure; unrelated." + "reason": "Local dataset loading in an image-classification script is unrelated to multimodal model loading failure for Phi-4. Different surfaces and root causes." }, { "left": "issue:46693", "right": "issue:47217", "accept": false, - "reason": "Long-context FA2 performance regression vs VL image-placeholder placement failure; unrelated." + "reason": "Performance regression with FA2 long-context generation is not the same as VL image-placeholder-in-user-text failures. Shared broad area only." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 import parsing error vs flash_attention s_aux=None crash; unrelated." + "reason": "Python 3.13 IndentationError from torch.jit parsing is unrelated to flash_attention.py crashing on s_aux=None. Completely different code paths." }, { "left": "issue:29870", "right": "issue:35141", "accept": false, - "reason": "BART generate decoder mask bug vs output embedding reinitialization on resize; unrelated." + "reason": "BART generate() decoder_attention_mask bug and embedding reinitialization after resize are different model-generation issues with no common fix." }, { - "left": "issue:45072", - "right": "issue:46932", + "left": "issue:45632", + "right": "issue:46617", "accept": false, - "reason": "Both are dtype-related, but they affect different models and different code paths; too broad to be the same bug." + "reason": "Both involve trust_remote_code/local cache, but one is a cache-path collision for same leaf names and the other is FileNotFoundError loading from cache. Related area, not same bug." }, { "left": "issue:44869", "right": "issue:46832", "accept": false, - "reason": "Whisper timestamp decoding crash vs RT-DETR feature-level mismatch; unrelated." + "reason": "Whisper trailing replacement-character crash and RT-DETR feature-level mismatch are unrelated decoding vs model-forward issues." }, { - "left": "issue:45632", - "right": "issue:46617", + "left": "issue:45072", + "right": "issue:46932", "accept": false, - "reason": "Both touch trust_remote_code/cache handling, but one is a leaf-directory path collision and the other is a FileNotFoundError from cache loading; different bugs." + "reason": "Both mention dtype issues in multimodal models, but one is a CI inference mismatch and the other is quantized Gemma4 vision/audio embedder casting. Not the same code path." }, { "left": "issue:45245", "right": "issue:46858", "accept": false, - "reason": "Category-count limit error vs static-cache sequence-length tensor arithmetic; unrelated." + "reason": "Category-count limit error and device-tensor seq_length arithmetic bug are unrelated. No shared underlying defect." }, { - "left": "issue:44734", - "right": "issue:45406", + "left": "issue:46082", + "right": "issue:46697", "accept": false, - "reason": "KV-cache continuation tensor indexing vs Gemma4Processor missing _tokenizer; unrelated." + "reason": "LlamaConfig head_dim validation and Swin2SR over-padding are different config vs preprocessing bugs." }, { - "left": "issue:46082", - "right": "issue:46697", + "left": "issue:44734", + "right": "issue:45406", "accept": false, - "reason": "LlamaConfig head_dim validation vs Swin2SR over-padding; unrelated." + "reason": "KV cache continuation indexing bug in serve is unrelated to Gemma4Processor missing _tokenizer. Different components and symptoms." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs still referencing removed pipeline() vs a runtime SigLIP2 pipeline/AutoModel issue; not the same change or bug." + "reason": "Docs still referencing pipeline() after removals is documentation cleanup, while SigLIP2 AutoModel/pipeline nonsense output is a runtime model/processor issue. Not duplicates." }, { - "left": "issue:44617", - "right": "issue:45584", + "left": "issue:43644", + "right": "issue:44913", "accept": false, - "reason": "Sam3Video OOM vs Whisper empty-transcription crash; unrelated." + "reason": "Non-persistent buffers filled with junk is not the same as rotary_pct silently resetting on config reload. Related to serialization only at a high level." }, { - "left": "issue:43644", - "right": "issue:44913", + "left": "issue:44617", + "right": "issue:45584", "accept": false, - "reason": "Non-persistent buffer initialization bug vs GPTNeoX config reload regression; unrelated." + "reason": "Sam3Video OOM and Whisper empty-transcription failure are unrelated. One is resource usage, the other is a generation edge case." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "Pipeline top_k nesting bug vs XLMRoberta FlashAttention 2 support issue; unrelated." + "reason": "top_k nested-list output bug in a pipeline and XLMRoberta Flash Attention 2 support issue are unrelated." }, { "left": "issue:43335", "right": "issue:44855", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug vs DebertaV2 import parsing error; unrelated." + "reason": "Sparse layer creation bug in SwitchTransformersConfig and a Python 3.13 IndentationError from torch.jit parsing are unrelated." }, { "left": "issue:43931", "right": "issue:44060", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch vs tied-weights warning misbinding dt_bias; different model bugs." + "reason": "Qwen3-VL weight-shape mismatch on load and Qwen3-Next tied-weights warning are different loading/config issues with different fixes." }, { - "left": "issue:44242", - "right": "issue:46897", + "left": "issue:44077", + "right": "issue:44877", "accept": false, - "reason": "MoE load-balancing loss omission vs Florence2 training-loss double-shift bug; unrelated." + "reason": "PatchTSMixer post_init policy and strict config blocking granite_speech are unrelated config-validation concerns." }, { - "left": "issue:44077", - "right": "issue:44877", + "left": "issue:44242", + "right": "issue:46897", "accept": false, - "reason": "PatchTSMixer post_init policy vs strict loading of granite_speech config; unrelated." + "reason": "Load-balancing loss not added when output_router_logits=False is not the same as Florence2 double-shift training-loss bug. Both are loss-related but different defects." }, { "left": "issue:43673", "right": "issue:46421", "accept": false, - "reason": "Both are cache-related, but one is missing cache in chunked prefill while the other is spurious recompilation from lazy cache init; different bugs." + "reason": "Both concern cache/chunked prefill, but one is missing GenerationMixin cache during chunked_prefill and the other is lazy cache initialization causing recompilation. Different symptoms and likely fixes." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "Paged generate warning vs Whisper empty-transcription failure; unrelated." + "reason": "Paged generate warning about num_return_sequences and Whisper empty-transcription failure are unrelated generation issues." } ] }, @@ -11992,11 +11984,11 @@ "issue:41720|issue:45245", "issue:44933|issue:45362", "issue:43335|issue:44190", - "issue:45588|issue:47013", "issue:44479|issue:44734", "issue:43065|issue:44610", - "issue:44617|issue:46369", "issue:43854|issue:44792", + "issue:45588|issue:47013", + "issue:44617|issue:46369", "issue:45381|issue:46655", "issue:39692|issue:44190", "issue:43295|issue:45276", @@ -12009,215 +12001,221 @@ "issue:25251|issue:29870", "issue:45290|issue:46752", "issue:44462|issue:45907", - "issue:43643|issue:44496", "issue:44521|issue:45381", + "issue:43643|issue:44496", "issue:44938|issue:45542", "issue:43854|issue:44964", "issue:43526|issue:46697", "issue:43673|issue:46858" ], - "prepared_review_unit_hash": "15745804c81a175296e17f5f26bad3ac12109332d5bd460e570736567c2b082e", + "prepared_review_unit_hash": "eb581171d38e55b0ce0a43621c489a910d7bbc41448539d44d285604fca3ff70", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36277, - "estimated_input_tokens": 9070, - "estimated_eval_tokens": 18396 + "serialized_chars": 36282, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue set with only one plausible duplicate edge: the two DeepSeek tokenizer regressions. The rest are superficial similarities across different models, processors, or failure modes and should stay separate.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 44779 is the broader original report of the DeepSeek tokenizer regression, and 46489 reads like a follow-up describing a gap in that fix.", + "summary": "Mostly heterogeneous issue cluster with no PRs; nearly all pairs are unrelated. The only plausible duplicate-like pair is the DeepSeek tokenizer regression pair (44779/46489).", + "confidence": 0.87, + "canonical_issue_reason": "issue:44779 is the broader, more general report of the DeepSeek tokenizer regression; issue:46489 looks like a more specific instance/gap of that same problem.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44521 is the most active and broadly impactful report in the set, with multiple inbound references and a concrete user-facing multimodal bug.", + "best_issue_reason": "issue:44779 is the best anchor for the cluster because it matches the only strong duplicate-like counterpart and describes the broader tokenizer regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45381", "right": "issue:46369", "accept": false, - "reason": "Different models and failures: Qwen2.5-VL video position IDs vs Sam3-LiteText memory overflow." + "reason": "Different models and failures: qwen2.5-vl video vision_position_ids vs Sam3-LiteText memory overflow." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "Unrelated areas: torch.compile/SDPA crash in Bamba vs local dataset loading in an example script." + "reason": "Unrelated subsystems: Bamba torch.compile/SDPA failure vs local dataset loading in an example script." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Different bugs: Qwen3 auto device mapping CUDA assert vs category-count limit in a processor." + "reason": "Different bug classes: Qwen3 auto device map CUDA assert vs category-count limit error." }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "Separate problems: a bad import from image_utils vs a Qwen3.5 chat crash." + "reason": "Unrelated issues: missing image_utils import vs Qwen3.5 chat crash." }, { "left": "issue:43335", "right": "issue:44190", "accept": false, - "reason": "SwitchTransformers config bug and local dataset loading are unrelated." - }, - { - "left": "issue:45588", - "right": "issue:47013", - "accept": false, - "reason": "Different code paths and symptoms: flash_attention sink handling vs xLSTM state unpacking." + "reason": "SwitchTransformers sparse-layer config bug is unrelated to dataset-loading failure." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Both are generation-related but about different regressions: Qwen video input vs serve KV-cache indexing." + "reason": "Both are regressions, but one is video-input handling and the other is /v1/responses KV-cache indexing." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "Sam3PixelDecoder dummy conv and OmDet processor size mismatch are distinct model/processor issues." + "reason": "Different code paths: dummy Conv2d in Sam3PixelDecoder vs OmDet-Turbo processor/model size mismatch." }, { - "left": "issue:44617", - "right": "issue:46369", + "left": "issue:43854", + "right": "issue:44792", "accept": false, - "reason": "Both mention memory, but one is OOM in Sam3Video and the other is an overflow in Sam3-LiteText." + "reason": "Unrelated model-loading problem vs janus test failure." }, { - "left": "issue:43854", - "right": "issue:44792", + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Different failures in different layers: flash_attention s_aux=None vs xLSTM return shape unpacking." + }, + { + "left": "issue:44617", + "right": "issue:46369", "accept": false, - "reason": "Different models and failures: GLM-4.7 loading in tests vs Janus image generation test failure." + "reason": "Both mention memory issues, but they are different models and different symptoms (OOM vs overflow)." }, { "left": "issue:45381", "right": "issue:46655", "accept": false, - "reason": "Both involve video processing, but the bugs are different: vision_position_ids vs padding calculation." + "reason": "Different multimodal bugs: wrong vision_position_ids vs incorrect padding calculation in SmolVLM video processing." }, { "left": "issue:39692", "right": "issue:44190", "accept": false, - "reason": "SigLIP2 doc/example problems are unrelated to the local dataset-loading script issue." + "reason": "SigLIP2 docs/model mismatch and quantization issues are unrelated to the local-dataset script bug." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Different regressions: processor.tokenizer/image handling vs gemma4 embedding resize propagation." + "reason": "Different APIs and regressions: processor.tokenizer access vs resize_token_embeddings not updating gemma4 embeddings." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "LoRA/vocab merge collapse and Qwen3.5 save_pretrained visual-key corruption are not the same bug." + "reason": "LoRA/vocab collapse is unrelated to Qwen3.5 visual-encoder save_pretrained key regression." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Chunked-generation inconsistency and a stale warning in paged generate are different issues." + "reason": "Both involve generation, but one is incorrect outputs and the other is a stale warning; not the same bug." }, { "left": "issue:44779", "right": "issue:46489", "accept": true, - "reason": "Both report the same DeepSeek tokenizer regression in v5+; 46489 explicitly describes a gap in the earlier fix, so one PR could plausibly address both." + "reason": "Same underlying DeepSeek tokenizer regression in transformers v5+: both report incorrect tokenizer output, with 46489 describing a specific gap in the same fix." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Unrelated regressions: Gemma4 embedding resize behavior vs Whisper empty-transcription generation failure." + "reason": "Gemma4 embedding-resize bug is unrelated to Whisper empty-transcription failure after align_special_tokens." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Qwen3OmniMoe video-input validation and serve KV-cache indexing are different failures." + "reason": "Qwen3OmniMoe video ValueError and serve KV-cache indexing are different problems." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "Both mention CI, but they target different tests and different broken mappings/config references." + "reason": "Separate CI issues affecting different tests and modules." }, { "left": "issue:25251", "right": "issue:29870", "accept": false, - "reason": "Different bugs: pipeline top_k output shape vs Bart generate() unbound decoder_attention_mask." + "reason": "Pipeline top_k nesting bug and Bart generate() decoder_attention_mask UnboundLocalError are unrelated." }, { "left": "issue:45290", "right": "issue:46752", "accept": false, - "reason": "Both are apply_chat_template crashes, but on different malformed inputs and different failure modes." + "reason": "Same function, but different edge cases: tool-call assistant messages vs an empty conversation list." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "Tokenizer selection and list_repo_files revision forwarding are unrelated issues." + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to list_repo_files not forwarding revision." }, { - "left": "issue:43643", - "right": "issue:44496", + "left": "issue:44521", + "right": "issue:45381", "accept": false, - "reason": "AutoConfig trust_remote_code field loss and unrecognized-model/model_type loading errors are distinct config-path problems." + "reason": "Both are multimodal/template issues, but assistant_mask generation and vision_position_ids are distinct code paths." }, { - "left": "issue:44521", - "right": "issue:45381", + "left": "issue:43643", + "right": "issue:44496", "accept": false, - "reason": "Both are multimodal, but assistant_masks and video position IDs are separate bugs." + "reason": "Different config-loading failures: missing fields under trust_remote_code vs missing model_type in config.json." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Python 3.14 import failure and missing TensorFlow backend handling are unrelated." + "reason": "Python 3.14 import/load failure is unrelated to the tensorboard-without-TF backend error." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Different model-loading failures for different models; no shared concrete bug." + "reason": "Different model-loading failures for different models; not the same underlying bug." }, { "left": "issue:43526", "right": "issue:46697", "accept": false, - "reason": "Beit fast label reduction and Swin2SR over-padding are separate image-processor bugs." + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Swin2SR over-padding." }, { "left": "issue:43673", "right": "issue:46858", "accept": false, - "reason": "Both involve cache/generation, but they fail in different places and for different reasons." + "reason": "Both touch cache/prefill, but one is missing GenerationMixin cache in chunked_prefill and the other is device-tensor slice arithmetic on non-CUDA backends." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Flash-attention sink handling and xLSTM train-mode return unpacking are unrelated failures." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted soft edge, 44779 \u2194 46489, is well supported by the issue titles, and the rest of the pair rejections are appropriately cautious. The selection of 44521 as the strongest issue is also defensible from the activity/inbound-reference counts and the multimodal user-facing bug described in its title." + "feedback": "The cluster summary is grounded in the packet: there are no PRs, the items are mostly heterogeneous issues, and 44779/46489 is the only clearly duplicate-like pair. The rejected soft edges are conservative and the accepted DeepSeek tokenizer pair is well supported by the titles and descriptions." }, "evaluator_used": true, "retried": false, @@ -12248,6 +12246,7 @@ "issue:42175", "issue:42796", "issue:43295", + "issue:43381", "issue:43404", "issue:43526", "issue:43531", @@ -12274,9 +12273,9 @@ "issue:45216", "issue:45335", "issue:45446", + "issue:45593", "issue:45600", "issue:45632", - "issue:45759", "issue:45803", "issue:46082", "issue:46121", @@ -12293,8 +12292,8 @@ "issue:44964|issue:45600", "issue:44112|issue:44964", "issue:42175|issue:44279", - "issue:45632|issue:46234", "issue:44496|issue:47217", + "issue:45632|issue:46234", "issue:32090|issue:35141", "issue:42796|issue:44521", "issue:36246|issue:41628", @@ -12307,211 +12306,218 @@ "issue:41950|issue:44016", "issue:43295|issue:45335", "issue:45083|issue:46506", - "issue:43582|issue:46932", - "issue:44496|issue:45042", "issue:43976|issue:44987", + "issue:44496|issue:45042", + "issue:43582|issue:46932", "issue:43673|issue:44336", "issue:43526|issue:43901", "issue:44683|issue:45446", - "issue:44496|issue:45759", - "issue:44016|issue:44220" + "issue:43901|issue:44496", + "issue:44016|issue:44220", + "issue:43381|issue:45593" ], - "prepared_review_unit_hash": "9d61b00ff7b1ac41b76f66b85bacdbaf09f5ce3ec578b8394a540fe0ed98d728", + "prepared_review_unit_hash": "56147211bfb31b717a2a78a6cd489322d13ff0acf5a88a5519ca9b36a265f49e", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35650, - "estimated_input_tokens": 8913, - "estimated_eval_tokens": 18082 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36452, + "estimated_input_tokens": 9113, + "estimated_eval_tokens": 18482 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are mostly false-positive similarities: the paired issues share broad themes like imports, multimodal handling, caching, or model config, but they describe different concrete bugs and code paths. No pull requests are present.", - "confidence": 0.89, - "canonical_issue_reason": "If one issue must serve as the anchor, issue:44521 is the strongest representative: it is actively discussed, has inbound references, and has a concrete multimodal repro. That said, this batch does not look like a true duplicate cluster.", + "summary": "This cluster is heterogeneous: the soft-similarity pairs mostly share only broad subsystem terms (imports, cache, multimodal, flex_attention, etc.) and do not describe the same concrete bug or change. I would not merge any of the candidate pairs.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the best standalone issue in this set because it is current, specific, and likely to attract follow-up from related multimodal-template work.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Import performance vs unsafe sys.modules access are different bugs and different code paths." + "reason": "Both are transformer internals, but one is a slow import concern and the other is a sys.modules safety bug; different symptoms and code paths." }, { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "Flash Attention support for XLM-Roberta is unrelated to embedding reinitialization after resize_token_embeddings." + "reason": "XLMRoberta Flash Attention support and token-embedding reinitialization are unrelated model behaviors." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "CPU-bound relative positional encoding and a bare-except bug in FuyuBatchFeature are unrelated." + "reason": "XLNet CPU tensor creation and a bare-except error swallowing in FuyuBatchFeature are unrelated fixes." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "Model loading failure for Phi-4 multimodal and stale auto_mappings references are different issues." + "reason": "A Phi-4 multimodal loading failure and stale auto_mappings references to Sam3LiteText are different model/config issues." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "A stale CI test in GraniteSpeech is not the same bug as loading Phi-4 multimodal." + "reason": "CI test flakiness in GraniteSpeech does not match a multimodal model-loading regression." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Missing TensorFlow in a torch install is a specific packaging bug; the other is only a generic dependency complaint." + "reason": "Package-install backend dependency omission is not the same as a generic dependency issue report." }, { - "left": "issue:45632", - "right": "issue:46234", + "left": "issue:44496", + "right": "issue:47217", "accept": false, - "reason": "Cache-path collisions for trust_remote_code do not overlap with a missing import utility symbol." + "reason": "Unrecognized model/config loading and VL image-placeholder handling are distinct multimodal loading problems." }, { - "left": "issue:44496", - "right": "issue:47217", + "left": "issue:45632", + "right": "issue:46234", "accept": false, - "reason": "Unrecognized model config and image-placeholder handling in VL models are different failure modes." + "reason": "Cache-path collision for trust_remote_code and a missing import_utils helper are unrelated." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer GPU broadcast of NoneType and post_init reinitializing embeddings are unrelated." + "reason": "Trainer GPU broadcast TypeError and post_init embedding reinitialization are different bugs." }, { "left": "issue:42796", "right": "issue:44521", "accept": false, - "reason": "Tokenizer BPE-marker handling is not the same as assistant-mask generation for multimodal templates." + "reason": "Tokenizer BPE marker handling and multimodal assistant_mask generation are separate text-processing issues." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Missing Qwen2_5_VLImageProcessor and missing AutoImageProcessor are both import errors but for different symbols and fixes." + "reason": "Missing Qwen2.5 VL image processor export and missing AutoImageProcessor import are different missing-symbol problems." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "These are distinct missing-import problems in different modules." + "reason": "Different missing imports from different modules; no shared concrete failure." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "lm_head tying in Mistral3ForConditionalGeneration is a different bug from Qwen3.5 save_pretrained checkpoint correctness." + "reason": "Weight tying in Mistral3 generation and Qwen3.5 save_pretrained checkpoint correctness are unrelated model bugs." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText config references are unrelated." + "reason": "Qwen3-MoE sliding window behavior and removed Sam3LiteText config references are not the same issue." }, { "left": "issue:44295", "right": "issue:46506", "accept": false, - "reason": "Reading position_ids after buffer registration is unrelated to CPU/CUDA result divergence." + "reason": "Buffer registration behavior and CPU/CUDA output divergence are different failure modes." }, { "left": "issue:43673", "right": "issue:46424", "accept": false, - "reason": "Missing GenerationMixin cache in chunked_prefill and max_cache_length being honored only for quantized cache are different cache bugs." + "reason": "Chunked prefill cache missing and max_cache_length handling for quantized cache are separate cache-feature bugs." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "LlamaConfig head_dim validation and rope-parameter JSON list handling are separate validation issues." + "reason": "LlamaConfig head_dim validation and rope-parameter JSON list handling are unrelated config-validation issues." }, { "left": "issue:41950", "right": "issue:44016", "accept": false, - "reason": "A pipeline selecting image processors and a notebook syntax error are unrelated." + "reason": "Video-classification pipeline processor lookup and a notebook syntax error are unrelated." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Processor/tokenizer regression with images is not the same as decoder.embed_tokens not updating after resize_token_embeddings." + "reason": "Processor/tokenizer multimodal regression and t5gemma token resize not affecting decoder embeddings are different code paths." }, { "left": "issue:45083", "right": "issue:46506", "accept": false, - "reason": "A helper returning wrong feature-extract lengths is different from CPU/CUDA output differences." + "reason": "A helper-length computation bug in qwen3_omni_moe and CPU/CUDA divergence in PPDocLayoutV3 are unrelated." }, { - "left": "issue:43582", - "right": "issue:46932", + "left": "issue:43976", + "right": "issue:44987", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError and Gemma4 quantized-embedder dtype casting are unrelated." + "reason": "A Python version compatibility problem and a model-loading failure under transformers>=5.1.0 are not the same defect." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Unrecognized model config and PIL processors incorrectly requiring torchvision are different root causes." + "reason": "Model loading/config recognition and PIL image processors requiring torchvision are different multimodal issues." }, { - "left": "issue:43976", - "right": "issue:44987", + "left": "issue:43582", + "right": "issue:46932", "accept": false, - "reason": "Python version compatibility for Transformers 5.1.0 and loading a specific model are not the same bug." + "reason": "AppleSilicon TypeError in caching_allocator_warmup and Gemma4 dtype-casting fix are unrelated." }, { "left": "issue:43673", "right": "issue:44336", "accept": false, - "reason": "Cache missing during chunked_prefill has nothing to do with ANSI codes in loading_report." + "reason": "Cache-related GenerationMixin behavior and ANSI codes in loading_report are entirely different concerns." }, { "left": "issue:43526", "right": "issue:43901", "accept": false, - "reason": "A BeitImageProcessorFast label-reduction bug is not the same as outdated pipeline docs." + "reason": "BeitImageProcessorFast label reduction and docs mentioning return_all_scores are different API/behavior mismatches." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both mention flex_attention, but one is a torch>=2.9 compile failure and the other is a bad version check for AuxRequest import; not the same concrete fix." + "reason": "Both touch flex_attention, but one is a compiled flex_attention failure on torch>=2.9 and the other is an incorrect version check for AuxRequest import; not the same concrete bug." }, { - "left": "issue:44496", - "right": "issue:45759", + "left": "issue:43901", + "right": "issue:44496", "accept": false, - "reason": "A model-config loading error and missing text_config unwrapping for composite configs are distinct problems." + "reason": "Pipeline docs wording and an unrecognized model-loading error are unrelated." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax error and _torch_extract_fbank_features behavior are unrelated." + "reason": "Notebook syntax error and _torch_extract_fbank_features behavior are different issues." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated training/inference bugs." } ] }, @@ -12558,7 +12564,6 @@ "issue:44351", "issue:44368", "issue:44466", - "issue:44496", "issue:44779", "issue:44792", "issue:44933", @@ -12567,14 +12572,16 @@ "issue:45276", "issue:45278", "issue:45335", + "issue:45357", "issue:45375", "issue:45405", "issue:45584", "issue:45588", - "issue:45593", "issue:45812", + "issue:46082", "issue:46123", "issue:46369", + "issue:46399", "issue:46424", "issue:46515", "issue:46614", @@ -12584,7 +12591,6 @@ "issue:47217" ], "soft_pairs": [ - "issue:43381|issue:45593", "issue:42907|issue:43441", "issue:43650|issue:44220", "issue:44016|issue:44062", @@ -12595,11 +12601,10 @@ "issue:44279|issue:46515", "issue:43479|issue:45375", "issue:45335|issue:45584", - "issue:43901|issue:44496", "issue:43299|issue:45216", "issue:43901|issue:47217", - "issue:44327|issue:46752", "issue:46614|issue:47085", + "issue:44327|issue:46752", "issue:46123|issue:46369", "issue:44279|issue:45405", "issue:43295|issue:44779", @@ -12612,222 +12617,212 @@ "issue:43976|issue:45405", "issue:44779|issue:45812", "issue:43381|issue:47000", - "issue:44466|issue:45276" + "issue:44466|issue:45276", + "issue:46082|issue:46399", + "issue:43299|issue:45357" ], - "prepared_review_unit_hash": "f92e68655094f54a5a76eca0897ad88debcd16c88e7c380945e89cd19c30123e", + "prepared_review_unit_hash": "a4b832bcc08f7da60d30e1d4e678327cd98e878c4971c8e97cbab6801df3ef2f", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35723, - "estimated_input_tokens": 8931, - "estimated_eval_tokens": 18118 + "serialized_chars": 36332, + "estimated_input_tokens": 9083, + "estimated_eval_tokens": 18422 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of Transformers regressions, mostly model/tokenizer/import/load-path bugs plus a few docs/tests issues. The soft pairs are not the same underlying bug/change, so none should be merged.", - "confidence": 0.63, - "canonical_issue_reason": "issue:45020 is the broadest representative of the recurring theme here: recent-version regressions breaking model loading and remote-code models. It is not a true duplicate umbrella for all items, but it best matches the cluster\u2019s dominant failure mode.", + "summary": "The set is mostly a collection of unrelated transformer regressions, docs issues, and model-specific bugs; the soft edges are superficial matches on broad themes like \u201cimport error,\u201d \u201ctokenizer,\u201d or \u201cversion regression,\u201d not true duplicates.", + "confidence": 0.18, + "canonical_issue_reason": "Issue 45020 is the broadest umbrella in the set: it describes recent-version breakage affecting models using `remote_code`, which is the closest high-level representative of the cluster\u2019s general regression theme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best available issue to represent the cluster because it captures the broad version-regression pattern more than the narrower, model-specific or docs-only reports.", + "best_issue_reason": "45020 is the best available representative, but it is still too broad and not a true duplicate of the other issues; it should only be used as a loose umbrella, not a merged duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43381", - "right": "issue:45593", - "accept": false, - "reason": "Different bugs: eval-mode gradient checkpointing vs D-FINE auxiliary losses when denoising is off." - }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Both involve Ministral, but one is save/dequantization and the other is FlashAttention loading/runtime failure." + "reason": "Same model family, but different failures: saving dequantized checkpoints vs FlashAttention loading/runtime crash." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "Unrelated; one is a placeholder issue, the other is an audio feature-extraction bug." + "reason": "Unrelated issues: placeholder text vs an audio feature extraction bug." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Docs notebook syntax error vs tokenizer AddedToken keyword conflict; different code paths." + "reason": "Docs notebook syntax error is unrelated to the tokenizer `AddedToken` keyword-argument bug." }, { "left": "issue:44792", "right": "issue:46424", "accept": false, - "reason": "Failed model test vs cache configuration behavior; not the same issue." + "reason": "Different code paths: Janus image-generation test failure vs cache-length handling in quantized cache." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "One is a Python-version compatibility bug, the other is a generic dependency problem; different root causes." + "reason": "Generic Python-version compatibility issue vs a dependency problem; too broad to be the same bug." }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both mention FlashAttention, but the failures are different models/code paths and not one fixable bug." + "reason": "Both touch attention code, but one is a Ministral FlashAttention failure and the other is an integration crash on `s_aux=None`." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Both are import errors, but for different symbols/modules and different causes." + "reason": "Both are import errors, but for different missing symbols/modules and different underlying causes." }, { "left": "issue:44279", "right": "issue:46515", "accept": false, - "reason": "Environment/version incompatibility in general vs a specific torch 2.4.1 incompatibility; not the same defect." + "reason": "Different dependency incompatibilities: one is a transformers dependency issue, the other is specifically torch 2.4.1." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Different multimodal config bugs affecting different model families and fields." + "reason": "Both involve config handling, but they concern different model families and different missing/defaulted fields." }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "Resize-token-embeddings behavior in Gemma vs Whisper generation after align_special_tokens; unrelated." - }, - { - "left": "issue:43901", - "right": "issue:44496", - "accept": false, - "reason": "Docs mismatch for TextClassificationPipeline vs model config import failure; unrelated." + "reason": "Embedding resize propagation is unrelated to Whisper empty-transcription generation after token alignment." }, { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Both are Qwen-related regressions, but one is model loading and the other is save_pretrained checkpoint serialization." + "reason": "Both are Qwen3.5 regressions, but one is model loading and the other is incorrect checkpoint serialization." }, { "left": "issue:43901", "right": "issue:47217", "accept": false, - "reason": "Documentation mismatch vs a runtime VL placeholder crash; different issues." + "reason": "Docs mismatch for a pipeline option is unrelated to VL placeholder parsing failures." }, { - "left": "issue:44327", - "right": "issue:46752", + "left": "issue:46614", + "right": "issue:47085", "accept": false, - "reason": "Different pipeline crashes in different subsystems; no shared underlying bug." + "reason": "Stale phonemizer backend state and tokenizer thread-safety are different subsystems and bugs." }, { - "left": "issue:46614", - "right": "issue:47085", + "left": "issue:44327", + "right": "issue:46752", "accept": false, - "reason": "Tokenizer phonemizer state bug vs thread-safety regression; not the same change." + "reason": "Different edge-case crashes in different utilities; only the generic index-error shape overlaps." }, { "left": "issue:46123", "right": "issue:46369", "accept": false, - "reason": "MaskGenerationPipeline batch handling vs SAM memory overflow; unrelated failures." + "reason": "Batch-finalization bug in generation results vs memory overflow in a specific model call." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "Both are dependency/version-related, but one is a generic dependency issue and the other is an unreleased minimum PEFT version bump." + "reason": "Both mention dependencies, but one is an incompatibility issue and the other is an unreleased PEFT version bump." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Both are v5 regressions around tokenization/processor behavior, but they affect different APIs and concrete failure modes." + "reason": "Both are tokenizer regressions, but they affect different models and different failure modes; not the same code-path bug." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Different documentation problems in different areas; not the same bug." + "reason": "Docs issues, but one is a SigLIP2 example/model mismatch and the other is outdated pipeline docs." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "SGD args ignored vs tie_word_embeddings warning; unrelated." + "reason": "Optimizer arguments being ignored is unrelated to a tied-embeddings warning during Qwen fine-tuning." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Different bugs: dtype propagation in Blip2 vs incorrect checkpoint serialization for Qwen3.5." + "reason": "One is dtype propagation on BLIP2 load, the other is Qwen3.5 checkpoint serialization." }, { "left": "issue:42796", "right": "issue:43650", "accept": false, - "reason": "Tokenizer BPE-marker handling vs a placeholder issue; no relation." + "reason": "Tokenizer marker handling and a placeholder issue titled \"ADD THE DATA\" are unrelated." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "Both are import-related, but 45278 is a broad umbrella and 44351 is a specific missing symbol; not the same concrete bug." + "reason": "A specific missing export (`HybridCache`) is not the same underlying bug as a broad wave of import errors after upgrade." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Video pipeline processor lookup vs remote_code model breakage; different subsystem and failure mode." + "reason": "Different pipeline bug vs broad remote_code regression; too far apart to merge as duplicates." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Python compatibility vs dependency floor bump; unrelated version problems." + "reason": "Python compatibility and unreleased PEFT pin are separate dependency problems." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they target different model families and produce different incorrect outputs." + "reason": "Both are tokenizer regressions, but one is Deepseek-specific incorrect results and the other is Granite token IDs; no clear shared bug." }, { "left": "issue:43381", "right": "issue:47000", "accept": false, - "reason": "Gradient checkpointing eval-mode issue vs RTDetr loss infeasibility; unrelated." + "reason": "Gradient-checkpointing eval-mode behavior is unrelated to RTDetr loss infeasible-cost-matrix failure." }, { "left": "issue:44466", "right": "issue:45276", "accept": false, - "reason": "Both touch tied-weights/embedding behavior, but one is save/serialization inconsistency and the other is resize_token_embeddings not propagating." + "reason": "Both involve tied/embedding-related model state, but one is serialization behavior and the other is `resize_token_embeddings` propagation." }, { - "left": "issue:43295", - "right": "issue:44779", + "left": "issue:46082", + "right": "issue:46399", "accept": false, - "reason": "Both are v5 tokenizer/processor regressions, but not the same concrete code-path problem." + "reason": "Llama config validation and Qwen3.5 kernel registration are different model/config bugs." }, { - "left": "issue:45020", - "right": "issue:45278", + "left": "issue:43299", + "right": "issue:45357", "accept": false, - "reason": "Broad regression reports, but one is remote_code model breakage and the other is import errors after upgrade; too different to merge." + "reason": "Both are Qwen3.5 regressions, but save/load serialization of different components is not the same underlying issue." } ] }, @@ -12853,6 +12848,7 @@ "nodes": [ "issue:33290", "issue:41950", + "issue:42175", "issue:42994", "issue:43299", "issue:43317", @@ -12865,6 +12861,7 @@ "issue:43854", "issue:43873", "issue:43931", + "issue:43976", "issue:44016", "issue:44242", "issue:44246", @@ -12879,7 +12876,6 @@ "issue:45081", "issue:45103", "issue:45200", - "issue:45357", "issue:45372", "issue:45397", "issue:45542", @@ -12892,7 +12888,6 @@ "issue:45907", "issue:46082", "issue:46209", - "issue:46399", "issue:46512", "issue:46851", "issue:47000", @@ -12900,8 +12895,6 @@ "issue:47217" ], "soft_pairs": [ - "issue:46082|issue:46399", - "issue:43299|issue:45357", "issue:43653|issue:45593", "issue:46082|issue:46512", "issue:41950|issue:45823", @@ -12909,215 +12902,204 @@ "issue:43644|issue:47145", "issue:43650|issue:46209", "issue:43299|issue:43854", - "issue:43931|issue:47145", "issue:44462|issue:45081", + "issue:43931|issue:47145", "issue:44683|issue:45542", - "issue:45588|issue:47000", "issue:44016|issue:44291", + "issue:45588|issue:47000", "issue:43479|issue:43525", "issue:45081|issue:45907", + "issue:44336|issue:46851", "issue:42994|issue:43317", "issue:33290|issue:45656", - "issue:44336|issue:46851", "issue:43441|issue:45834", "issue:44623|issue:44749", "issue:43873|issue:47145", + "issue:44261|issue:44263", "issue:45859|issue:46851", "issue:45103|issue:45372", - "issue:44261|issue:44263", "issue:44246|issue:47217", - "issue:44623|issue:45397" + "issue:44623|issue:45397", + "issue:42175|issue:43976" ], - "prepared_review_unit_hash": "30574cb8bbb5be30ce77db46daf4ba60c91c0232a866fd9ed35837fc67f4b595", + "prepared_review_unit_hash": "ba2209d00ad7c9f3aa4c03e3ab782760622374f269764f193cac89b76ab89f03", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35199, - "estimated_input_tokens": 8800, - "estimated_eval_tokens": 17856 + "soft_pair_count": 26, + "serialized_chars": 34840, + "estimated_input_tokens": 8710, + "estimated_eval_tokens": 17676 }, "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; all candidates should stay separate. The strongest standalone issue in this set is the open quantization/offload report, but it is not a duplicate of the other quantization-related items.", - "confidence": 0.87, - "canonical_issue_reason": "issue:43873 is the broadest, most actively discussed open regression here and is the best anchor for quantization/offload triage, even though it is not a duplicate of the other issues.", - "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the most suitable representative issue overall because it is open, has substantial discussion/inbound references, and describes a concrete, high-impact inference/offloading failure.", - "best_pr_reason": null, + "summary": "This cluster is highly heterogeneous: the items cover unrelated tokenizer, multimodal config, quantization/offload, deepspeed, model-loading, and docs/notebook bugs. None of the soft pairs look like the same underlying defect, so no duplicate merge is justified.", + "confidence": 0.95, + "canonical_issue_reason": "No canonical issue stands out because the cluster does not share one underlying bug or change; the issues span many unrelated subsystems and model families.", + "canonical_pr_reason": "No PRs are present, so there is no canonical PR.", + "best_issue_reason": "No single issue is a good global representative: any pick would be arbitrary given the lack of a common failure mode across the set.", + "best_pr_reason": "No PRs are present, so there is no best PR candidate.", "soft_edge_verdicts": [ - { - "left": "issue:46082", - "right": "issue:46399", - "accept": false, - "reason": "Different failures in different code paths: LlamaConfig validation vs Qwen3.5 rotary kernel registration." - }, - { - "left": "issue:43299", - "right": "issue:45357", - "accept": false, - "reason": "Both are Qwen-related, but one is a model load failure and the other is save_pretrained key serialization; not the same bug." - }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "Tokenizer special-token registration is unrelated to D-FINE auxiliary-loss behavior." + "reason": "BigBirdTokenizer special-token decoding and D-FINE auxiliary-loss behavior are unrelated subsystems with different symptoms." }, { "left": "issue:46082", "right": "issue:46512", "accept": false, - "reason": "Head_dim validation and dtype/config-staleness are separate config-management bugs." + "reason": "LlamaConfig head_dim validation and from_config(dtype) stale config.dtype are different config-handling bugs." }, { "left": "issue:41950", "right": "issue:45823", "accept": false, - "reason": "Pipeline processor lookup vs device_map/offload mismatch are unrelated." + "reason": "Pipeline processor lookup and Gemma4 device-map forward mismatch are not the same code-path problem." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Loss gating on router logits is not the same as a multimodal token-id defaulting issue." + "reason": "Missing load-balancing loss and Gemma4 mm_token_type_ids defaults are separate model-training issues." }, { "left": "issue:43644", "right": "issue:47145", "accept": false, - "reason": "Both involve loading/corruption concerns, but one is non-persistent buffer initialization and the other is FP8 weight_scale application; different root causes." + "reason": "Both involve loading/state handling, but one is non-persistent buffers and the other is FP8 weight_scale application; different root causes." }, { "left": "issue:43650", "right": "issue:46209", "accept": false, - "reason": "Placeholder title is not indicative of the FP8/B200 inference bug." + "reason": "Placeholder title aside, this is unrelated to FP8 inference on B200." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Different models and different load-time failures." + "reason": "Different model families and different loading failures; not one shared bug." }, { - "left": "issue:43931", - "right": "issue:47145", + "left": "issue:44462", + "right": "issue:45081", "accept": false, - "reason": "Shape-mismatch loading of Qwen3-VL is not the same as FP8 frozen-checkpoint weight scaling." + "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashes are separate tokenizer-loading paths." }, { - "left": "issue:44462", - "right": "issue:45081", + "left": "issue:43931", + "right": "issue:47145", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashes are distinct tokenizer-loading bugs." + "reason": "Weight-shape mismatch for Qwen3-VL and FP8 frozen-checkpoint scaling are unrelated." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Compiled flex_attention/torch compatibility has nothing to do with tensorboard-only TensorFlow backend handling." + "reason": "Compiled flex_attention on torch 2.9 and tensorboard-without-TensorFlow backend errors have no shared underlying defect." }, { - "left": "issue:45588", - "right": "issue:47000", + "left": "issue:44016", + "right": "issue:44291", "accept": false, - "reason": "FlashAttention s_aux handling and RTDetr loss infeasibility are unrelated." + "reason": "Notebook syntax error and init_empty_weights argument mismatch are unrelated." }, { - "left": "issue:44016", - "right": "issue:44291", + "left": "issue:45588", + "right": "issue:47000", "accept": false, - "reason": "Notebook syntax error vs init_empty_weights argument mismatch are entirely different issues." + "reason": "Flash-attention AttributeError on s_aux=None and RTDetrLoss infeasible cost matrix are different model/runtime bugs." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Different config attributes and different model families." + "reason": "Both are config issues, but they affect different models and different fields; not the same bug." }, { "left": "issue:45081", "right": "issue:45907", "accept": false, - "reason": "Tokenizer regex crash and list_repo_files revision forwarding are unrelated." + "reason": "Mistral tokenizer regex patch crash and list_repo_files revision forwarding are unrelated." + }, + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI escape emission in loading_report and Qwen3.5 variable-length position_ids are unrelated." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Saving quantized models and loading dequantized models with offload are different operations and failures." + "reason": "Quantized model saving failure and dequantized-model offload loading failure are distinct issues." }, { "left": "issue:33290", "right": "issue:45656", "accept": false, - "reason": "Both mention deepspeed, but one is Adafactor OOM and the other is double optimizer-step invocation." - }, - { - "left": "issue:44336", - "right": "issue:46851", - "accept": false, - "reason": "ANSI color output in non-terminal and variable-length position_ids handling are unrelated." + "reason": "Deepspeed Adafactor OOM and deepspeed optimizer.step being called twice are different bugs despite the shared framework." }, { "left": "issue:43441", "right": "issue:45834", "accept": false, - "reason": "FlashAttention failure for Ministral-3 and OCR index error in Kosmos2.5 are different model-specific bugs." + "reason": "Ministral FlashAttention failure and Kosmos2.5 long-OCR index error are unrelated." }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "processor.save_pretrained file omission is not the same as the filtering slowdown report." + "reason": "processor.save_pretrained missing files and import slowdown after upgrade are unrelated." }, { "left": "issue:43873", "right": "issue:47145", "accept": false, - "reason": "Both are quantization/offload adjacent, but one is offloading behavior and the other is FP8 frozen-checkpoint weight_scale omission; not mergeable as one fix." + "reason": "General quantization offloading behavior and FP8 weight_scale omission are different problems." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Both are GLM-related, but one is a missing rms_norm_eps default and the other is a torch.split indexing issue." }, { "left": "issue:45859", "right": "issue:46851", "accept": false, - "reason": "CPU offload incompatibility in rotary embedding and variable-length position_ids behavior are different bugs." + "reason": "CPU offload incompatibility in a Qwen3.5 rotary embedding and variable-length position_ids are not the same failure mode." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "auto_docstring annotation handling is unrelated to Gemma 4 processor import failure." - }, - { - "left": "issue:44261", - "right": "issue:44263", - "accept": false, - "reason": "Layernorm epsilon mismatch and torch.split return-value handling are separate issues." + "reason": "Auto-docstring annotation handling and Gemma 4 processor import failure are unrelated." }, { "left": "issue:44246", "right": "issue:47217", "accept": false, - "reason": "Import latency is unrelated to vision-language placeholder placement failures." + "reason": "Import-time slowness and VL image-placeholder handling are unrelated." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Processor save_pretrained missing files and gemma-4 zero3 from_pretrained are different model-loading/saving bugs." + "reason": "Missing processor save files and gemma-4 zero3 from_pretrained are distinct issues." } ] }, @@ -13142,10 +13124,10 @@ "cluster_id": "cluster-10000-517", "nodes": [ "issue:41628", - "issue:42175", "issue:42796", "issue:43441", "issue:43525", + "issue:43526", "issue:43575", "issue:43643", "issue:43673", @@ -13191,238 +13173,238 @@ "issue:47217" ], "soft_pairs": [ - "issue:42175|issue:43976", "issue:47176|issue:47177", "issue:45600|issue:45712", "issue:44671|issue:44779", "issue:43525|issue:43881", "issue:43901|issue:44246", "issue:43673|issue:43761", + "issue:44188|issue:44336", "issue:43575|issue:47145", "issue:44016|issue:45823", - "issue:44188|issue:44336", "issue:44186|issue:47000", - "issue:43525|issue:43643", - "issue:44987|issue:45542", "issue:43824|issue:43976", + "issue:44987|issue:45542", + "issue:43525|issue:43643", "issue:44336|issue:44373", "issue:43723|issue:44162", "issue:42796|issue:43441", "issue:44589|issue:45362", "issue:45325|issue:46851", "issue:46291|issue:46399", - "issue:44246|issue:44496", "issue:44297|issue:45397", - "issue:44871|issue:47145", + "issue:44246|issue:44496", "issue:45083|issue:46851", + "issue:44871|issue:47145", "issue:41628|issue:43976", - "issue:43784|issue:45542", "issue:46655|issue:46726", + "issue:43784|issue:45542", "issue:45003|issue:47217", - "issue:46399|issue:46851" + "issue:46399|issue:46851", + "issue:43526|issue:45003" ], - "prepared_review_unit_hash": "57453059a357525b118af553d68e4b5c41938eeba5a38d6f92550cd2dad7029d", + "prepared_review_unit_hash": "3abc04b3ac4d9544c83ffc68af38a1f58e2576a6d5c568090891c8a3892cb41c", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36312, - "estimated_input_tokens": 9078, - "estimated_eval_tokens": 18412 + "serialized_chars": 36341, + "estimated_input_tokens": 9086, + "estimated_eval_tokens": 18428 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the items span unrelated import, tokenizer, config, kernel, packaging, and security issues. None of the soft-similarity pairs look like true duplicates or a single fixable change, so there is no safe canonical issue/PR here.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is a loose grab-bag of mostly unrelated Transformers issues: model/tokenizer loading regressions, config mismatches, processor bugs, kernel/runtime failures, docs typos, and a couple security reports. There is no strong duplicate center.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue is a good canonical representative because the items are too heterogeneous and do not describe one underlying bug.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No issue is a true global best fit; if one had to be picked, issue:43723 is the broadest user-facing loading regression, but it still does not unify the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42175", - "right": "issue:43976", - "accept": false, - "reason": "Different problems: backend extras installation vs Python version compatibility." - }, { "left": "issue:47176", "right": "issue:47177", "accept": false, - "reason": "Both are security-related, but one is command-substitution allowlist bypass and the other is archive path traversal/RCE." + "reason": "Different security bugs: command-substitution allowlist bypass vs tar/zip slip arbitrary file write." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Different maintenance failures: removed config references in CI vs leftover dummy PT classes." + "reason": "Both are repo hygiene problems, but one is stale auto_mappings config refs and the other is dummy classes leaking into __dir__/check_repo." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Different models and bug classes: CamemBERT masked-LM output vs DeepSeek tokenizer formatting." + "reason": "Different models and failures: CamemBERT MLM predictions vs Deepseek tokenizer output corruption." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Different model load failures; only the general symptom overlaps." + "reason": "Different loading failures with different root causes; only the symptom is a generic model-load crash." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Docs mismatch for a pipeline flag vs intermittent import slowdown." + "reason": "Docs mismatch for return_all_scores is unrelated to intermittent import slowness." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Different code paths: generation cache regression vs CLIP vision hidden_states regression." + "reason": "Both are v5 regressions, but one is chunked_prefill cache handling and the other is CLIPVision hidden_states propagation." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel branching under torch.compile is unrelated to ANSI escapes in loading_report." }, { "left": "issue:43575", "right": "issue:47145", "accept": false, - "reason": "One is OOM during tensor-parallel load; the other is silent FP8 weight-scale application corruption." + "reason": "Tensor-parallel OOM while loading Qwen2-57B is unrelated to FP8 frozen-checkpoint weight_scale handling." }, { "left": "issue:44016", "right": "issue:45823", "accept": false, - "reason": "Notebook syntax error vs model forward device mismatch." - }, - { - "left": "issue:44188", - "right": "issue:44336", - "accept": false, - "reason": "Attention-kernel divergence under torch.compile vs ANSI leakage in loading report." + "reason": "Notebook syntax error and Gemma4 forward device mismatch are different problems." }, { "left": "issue:44186", "right": "issue:47000", "accept": false, - "reason": "Tokenizer padding/truncation crash vs loss assignment infeasibility in RTDetr." + "reason": "LayoutLMv2 tokenizer padding/truncation crash is unrelated to RTDetrLoss cost-matrix infeasibility." }, { - "left": "issue:43525", - "right": "issue:43643", + "left": "issue:43824", + "right": "issue:43976", "accept": false, - "reason": "Both mention missing config fields, but they are distinct models and distinct failure modes." + "reason": "Qwen2_5_VL import error and Python version compatibility failure are separate issues." }, { "left": "issue:44987", "right": "issue:45542", "accept": false, - "reason": "Different dependency/backend issues: loading a model package vs missing TensorFlow backend when only tensorboard is installed." + "reason": "Model loading failure on v5.1.0 and missing TensorFlow backend when only tensorboard is installed are unrelated." }, { - "left": "issue:43824", - "right": "issue:43976", + "left": "issue:43525", + "right": "issue:43643", "accept": false, - "reason": "Import error for a specific VL class vs package Python-version support regression." + "reason": "Both mention missing fields, but one is a specific Llama4Config attribute bug and the other is trust_remote_code AutoConfig field loss." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "Unrelated: terminal ANSI formatting bug vs wrong docstring." + "reason": "ANSI code leakage in loading_report is unrelated to a wrong docstring." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer-loading regression vs ESM2 model breakage." + "reason": "Tokenizer loading regression and ESM2 breakage are different model-specific failures." }, { "left": "issue:42796", "right": "issue:43441", "accept": false, - "reason": "Different tokenizer bug and FlashAttention/model-loading regression." + "reason": "Ministral tokenizer BPE-marker handling and FlashAttention failure are distinct bugs." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage lookup failure vs chat/variable-length Qwen3.5 crash." + "reason": "Float8 storage lookup failure is unrelated to Qwen3.5 chat loading/crash behavior." }, { "left": "issue:45325", "right": "issue:46851", "accept": false, - "reason": "Both involve position_ids, but the affected models and specific bugs differ." + "reason": "Both concern position_ids, but they affect different model families and different logic paths." }, { "left": "issue:46291", "right": "issue:46399", "accept": false, - "reason": "Both mention kernels, but one is import-time dependency validation and the other is a registration/decorator crash." - }, - { - "left": "issue:44246", - "right": "issue:44496", - "accept": false, - "reason": "Import slowness is unrelated to unrecognized model config handling." + "reason": "kernels import/version requirement failure is unrelated to a missing decorator on apply_rotary_pos_emb registration." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer save metadata mismatch vs Gemma-4 zero3 from_pretrained issue." + "reason": "Tokenizer config class mismatch is not the same bug as gemma-4 zero3 from_pretrained failure." }, { - "left": "issue:44871", - "right": "issue:47145", + "left": "issue:44246", + "right": "issue:44496", "accept": false, - "reason": "EOS-token config inconsistency is unrelated to FP8 checkpoint scale application." + "reason": "Import latency and unrecognized model config are unrelated." }, { "left": "issue:45083", "right": "issue:46851", "accept": false, - "reason": "Different helper/output-length logic vs variable-length position_ids regression." + "reason": "Different Qwen-family position handling bugs in different code paths." }, { - "left": "issue:41628", - "right": "issue:43976", + "left": "issue:44871", + "right": "issue:47145", "accept": false, - "reason": "Import name exposure issue vs Python version support problem." + "reason": "Gemma eos_token_id inconsistency is unrelated to FP8 weight_scale loading." }, { - "left": "issue:43784", - "right": "issue:45542", + "left": "issue:41628", + "right": "issue:43976", "accept": false, - "reason": "Import NameError in sentence-transformers integration vs missing TensorFlow backend packaging." + "reason": "AutoImageProcessor import failure and Python 5.1.0 compatibility are not the same issue." }, { "left": "issue:46655", "right": "issue:46726", "accept": false, - "reason": "Both are processor-related, but one is padding math in video processing and the other is missing-vision-backend error handling." + "reason": "Video padding miscalculation and missing-vision-backend image processor error are different processor problems." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "sentence-transformers import NameError and undefined tf backend error have different causes." }, { "left": "issue:45003", "right": "issue:47217", "accept": false, - "reason": "Unrelated internals access bug vs VL image-placeholder handling failure." + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to VL placeholder handling in user text." }, { "left": "issue:46399", "right": "issue:46851", "accept": false, - "reason": "Different Qwen3.5 failures: kernel registration crash vs variable-length position_ids bug." + "reason": "Kernel registration crash and variable-length position_ids bug are different Qwen3.5 issues." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to modeling_utils sys.modules access." } ] }, @@ -13453,7 +13435,6 @@ "issue:42886", "issue:43257", "issue:43299", - "issue:43526", "issue:43531", "issue:43540", "issue:43550", @@ -13467,6 +13448,7 @@ "issue:44451", "issue:44479", "issue:44485", + "issue:44749", "issue:44779", "issue:44871", "issue:45003", @@ -13496,7 +13478,6 @@ "issue:47217" ], "soft_pairs": [ - "issue:43526|issue:45003", "issue:43540|issue:46752", "issue:45103|issue:46399", "issue:45072|issue:46829", @@ -13504,8 +13485,8 @@ "issue:44295|issue:45397", "issue:43531|issue:45362", "issue:44246|issue:44871", - "issue:45588|issue:46851", "issue:44336|issue:44485", + "issue:45588|issue:46851", "issue:46614|issue:46665", "issue:42491|issue:43299", "issue:45003|issue:46821", @@ -13522,198 +13503,199 @@ "issue:43540|issue:46655", "issue:44451|issue:46752", "issue:45375|issue:45759", - "issue:44479|issue:46655" + "issue:44479|issue:46655", + "issue:44295|issue:44749" ], - "prepared_review_unit_hash": "297e8beb440b4a62e6f4416105a7685f22eb12d95ce27bce448600ff121625c0", + "prepared_review_unit_hash": "5c8ca2f0553b227f3e233ce5596fc0d113cf13d74db878b5b9a048b97bb8a6f3", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35744, - "estimated_input_tokens": 8936, - "estimated_eval_tokens": 18128 + "serialized_chars": 35768, + "estimated_input_tokens": 8942, + "estimated_eval_tokens": 18140 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous, but the closest recurring theme is Qwen/VL/MoE loading and video-input regressions. None of the suggested soft pairs look like true duplicates; they mostly hit different models, different pipeline stages, or unrelated utilities.", - "confidence": 0.89, - "canonical_issue_reason": "issue:44479 is the broadest, clearest representative of the recurring Qwen video-input regression theme, covering multiple affected Qwen VL/MoE variants in one concrete bug report.", + "summary": "This cluster is mostly heterogeneous: several Qwen/VL/video/loading regressions are mixed with unrelated import, docs, performance, and training issues. I did not find any pair that clearly looks like the same underlying bug, so all soft edges should be rejected. ", + "confidence": 0.67, + "canonical_issue_reason": "issue:44479 is the broadest and most explicit regression report in the set, spanning multiple Qwen VL/MoE variants with a clear version boundary (v5.3.0) and a concrete failure mode. If one issue must anchor this mixed cluster, it is the best representative.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the strongest single issue to anchor this cluster because it most directly captures the shared user-facing regression pattern across the related Qwen video/VL reports.", + "best_issue_reason": "issue:44479 is the most reusable cluster representative because it describes a concrete, multi-model regression rather than a narrow one-off symptom. The rest of the items are either unrelated or too specific to serve as a stable umbrella.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43526", - "right": "issue:45003", - "accept": false, - "reason": "Different subsystems: BEiT image-processor label reduction vs unsafe sys.modules access in modeling_utils." - }, { "left": "issue:43540", "right": "issue:46752", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError and empty-chat-template IndexError are unrelated code paths." + "reason": "Both are runtime failures, but one is Qwen3OmniMoe video-input handling and the other is empty-chat-template handling; different code paths and causes." }, { "left": "issue:45103", "right": "issue:46399", "accept": false, - "reason": "auto_docstring annotation parsing bug is unrelated to Qwen3.5 kernel registration/decorator failure." + "reason": "One is an auto-docstring annotation crash, the other is a Qwen3.5 kernel-registration error; unrelated subsystems." }, { "left": "issue:45072", "right": "issue:46829", "accept": false, - "reason": "Distinct model/output bugs: dtype mismatch in SwitchTransformers/TimmWrapperModel vs hidden_states field mixup in VideoPrism classification output." + "reason": "bfloat16 dtype mismatch in inference vs incorrect hidden_states assignment in a video-classification output; not the same bug." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "pytest-xdist file-race in testing utils is unrelated to the stale num_return_sequences warning in paged generate()." + "reason": "pytest-xdist file race and stale paged-generate warning are unrelated behavior issues." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "Buffer access on position_ids and Gemma-4 ZeRO3 loading failure are different failure modes." + "reason": "position_ids buffer access vs gemma-4 zero3 from_pretrained failure; different failure mechanisms." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is not the same bug as Qwen3.5-35B chat crash." + "reason": "Qwen3-MoE sliding_window handling and Qwen3.5 chat crash are distinct model-path problems." }, { - "left": "issue:44246", + "left": "issue:44242", "right": "issue:44871", "accept": false, - "reason": "Load-balancing loss omission and Gemma eos_token_id configuration mismatch are unrelated." + "reason": "Missing load-balancing loss and inconsistent eos_token_id config affect different parts of different models." }, { - "left": "issue:45588", - "right": "issue:46851", + "left": "issue:44336", + "right": "issue:44485", "accept": false, - "reason": "flash_attention s_aux=None crash and variable-length position_ids bug affect different paths." + "reason": "Terminal-color output bug in loading_report vs GLM-5 RoPE implementation discussion; no shared underlying defect." }, { - "left": "issue:44336", - "right": "issue:44485", + "left": "issue:45588", + "right": "issue:46851", "accept": false, - "reason": "ANSI escape handling in loading_report is unrelated to GLM-5 RoPE implementation." + "reason": "Flash-attention sinkless-model crash and variable-length position_ids issue are separate runtime bugs." }, { "left": "issue:46614", "right": "issue:46665", "accept": false, - "reason": "Stale phonemizer backend state is a different bug from inconsistent from_pretrained behavior inside classes." + "reason": "Stale phonemizer backend state vs from_pretrained behavior inside Python classes; unrelated." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both involve Qwen3 MOE, but one is adapter/accelerate-deepspeed loading and the other is Qwen3VL loading on v5.0.dev0; not the same concrete bug." + "reason": "Both involve Qwen3 MoE, but one is a conversion/deepspeed issue and the other is a loading regression in v5; not the same concrete bug." }, { "left": "issue:45003", "right": "issue:46821", "accept": false, - "reason": "Unsafe sys.modules access is unrelated to TimesFM window_size attribute errors." + "reason": "Unsafe sys.modules access and TimesFM window_size AttributeError are unrelated implementation bugs." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Router-logits loss handling and D-FINE auxiliary-loss behavior are different training bugs." + "reason": "Both mention losses, but they are different models and different missing-loss conditions." }, { "left": "issue:43550", "right": "issue:46506", "accept": false, - "reason": "Bamba torch.compile/SDPA failure and PPDocLayout CPU-vs-CUDA discrepancy are unrelated." + "reason": "torch.compile/SDPA failure for Bamba and CPU/CUDA output divergence for PPDocLayoutV3 are distinct issues." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline cache loading failure is not the same as AttentionInterface.register changing registered-function behavior." + "reason": "Offline cache loading and AttentionInterface.register behavior change are unrelated." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "ImportError for Qwen2_5_VLImageProcessor and missing is_offline_mode are separate import issues." + "reason": "Both are import errors, but for different symbols and different root causes." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "DeepSeek tokenizer regression and resize_token_embeddings/output_embeddings mismatch do not share a concrete code path." + "reason": "Tokenizer correctness regression and resize_token_embeddings not updating output embeddings are different bugs." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Different missing symbols: Qwen2_5_VLImageProcessor vs HybridCache." + "reason": "Different missing imports in different modules; same symptom class but not the same underlying issue." }, { "left": "issue:46726", "right": "issue:47217", "accept": false, - "reason": "Missing vision backend handling and image-placeholder-in-user-text failure are different VL issues." + "reason": "Missing vision backend handling vs image placeholder-in-text failure are separate VL issues." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "A docs/behavior mismatch in TextClassificationPipeline is unrelated to modeling_utils sys.modules access." + "reason": "Documentation mismatch and unsafe sys.modules access are unrelated." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and DeepSpeed ZeRO3 deque underflow are distinct failures." + "reason": "Both involve DeepSpeed, but one is model weight conversion and the other is a ZeRO3 deque error; not mergeable as one fix." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification pipeline processor lookup bug and _torch_extract_fbank_features() issue are unrelated." + "reason": "Video-classification pipeline image-processor lookup and fbank feature extraction are unrelated." }, { "left": "issue:43540", "right": "issue:46655", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError and SmolVLM padding calculation bug do not look like the same underlying defect." + "reason": "Both touch video processing, but they are different models and different defects (input parsing vs padding calculation)." }, { "left": "issue:44451", "right": "issue:46752", "accept": false, - "reason": "ScandiBERT model loading failure and empty-list chat_template crash are unrelated." + "reason": "ScandiBERT loading failure and empty-list chat template crash are unrelated." }, { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Missing deepstack_visual_indexes config field and text_config unwrapping in from_config are different composite-config issues." + "reason": "Missing config field under @strict and unwrapping text_config for composite multimodal configs are different loading/config bugs." }, { "left": "issue:44479", "right": "issue:46655", "accept": false, - "reason": "Both are video-related, but 44479 is a Qwen v5.3.0 regression across multiple models while 46655 is a SmolVLM padding-calculation bug." + "reason": "Both are video-related, but one is a broad Qwen VL regression and the other is a SmolVLM padding bug; different code paths." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Position_ids buffer error and a performance regression after upgrade are not the same defect." } ] }, @@ -13741,10 +13723,13 @@ "issue:30333", "issue:36246", "issue:42199", + "issue:43065", "issue:43295", "issue:43352", + "issue:43408", "issue:43475", "issue:43526", + "issue:43540", "issue:43596", "issue:43653", "issue:44038", @@ -13752,12 +13737,10 @@ "issue:44077", "issue:44162", "issue:44242", - "issue:44295", "issue:44368", "issue:44466", "issue:44479", "issue:44623", - "issue:44749", "issue:44857", "issue:44877", "issue:44977", @@ -13786,210 +13769,224 @@ "issue:47000" ], "soft_pairs": [ - "issue:44295|issue:44749", "issue:44162|issue:45200", - "issue:45859|issue:46399", "issue:45292|issue:47000", + "issue:45859|issue:46399", "issue:45137|issue:45600", - "issue:45003|issue:45561", "issue:43295|issue:45292", + "issue:45003|issue:45561", "issue:36246|issue:43596", "issue:43475|issue:46369", "issue:44466|issue:45335", "issue:44623|issue:46726", "issue:44038|issue:46506", - "issue:44242|issue:44857", - "issue:44877|issue:45375", "issue:43526|issue:46897", + "issue:44877|issue:45375", + "issue:44242|issue:44857", "issue:44987|issue:45459", "issue:29870|issue:30333", "issue:43352|issue:45250", "issue:43526|issue:46894", "issue:44479|issue:46170", "issue:42199|issue:46851", - "issue:44060|issue:44077", - "issue:43653|issue:44162", "issue:45397|issue:45656", - "issue:44368|issue:44977" + "issue:43653|issue:44162", + "issue:44368|issue:44977", + "issue:44060|issue:44077", + "issue:43065|issue:43408", + "issue:44877|issue:46170", + "issue:43475|issue:43540" ], - "prepared_review_unit_hash": "02f38d797534c362c178b52570bc930c858eee87748ab35322e07eb3ca5ec109", + "prepared_review_unit_hash": "91f959f1d121b7e903173fea5162cc4a6ca8e560bfe26af2de84333ab1bcae04", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 25, - "serialized_chars": 34816, - "estimated_input_tokens": 8704, - "estimated_eval_tokens": 17664 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35894, + "estimated_input_tokens": 8974, + "estimated_eval_tokens": 18204 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a true duplicate set: the items span unrelated bugs across many models and subsystems, and none of the soft pairs look like the same concrete issue. I would not merge any of the proposed soft edges.", - "confidence": 0.98, - "canonical_issue_reason": "No clear canonical issue: the items are heterogeneous and do not share one underlying bug or change.", + "summary": "This cluster is highly heterogeneous: most items are unrelated model-specific regressions, config issues, or infrastructure bugs. None of the soft pairs looks like the same underlying fix, so there is no strong duplicate hub.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue cleanly represents the cluster; the titles span unrelated bugs across different models, runtimes, and subsystems.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue candidate; the list mixes unrelated reports, so any single issue would be arbitrary.", + "best_issue_reason": "If a representative issue must be chosen, 44479 is the broadest user-facing regression in the set because it impacts several Qwen video variants; still, this is not a true duplicate anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44295", - "right": "issue:44749", - "accept": false, - "reason": "Both mention position_id/serialization behavior, but one is about reading a registered buffer and the other is about slowdown after upgrading Transformers; different failure modes and code paths." - }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 breakage and Gemma 4 mm_token_type_ids are unrelated model-specific bugs." + "reason": "Different models and failures: ESM2 is broken vs Gemma 4 needs mm_token_type_ids defaults." }, { - "left": "issue:45859", - "right": "issue:46399", + "left": "issue:45292", + "right": "issue:47000", "accept": false, - "reason": "Different Qwen3.5 MoE problems: one is CPU-offload incompatibility in rotary embedding, the other is a kernel registration crash on apply_rotary_pos_emb." + "reason": "Unrelated code paths: token embedding resize bug vs RTDetr loss infeasible cost matrix." }, { - "left": "issue:45292", - "right": "issue:47000", + "left": "issue:45859", + "right": "issue:46399", "accept": false, - "reason": "resize_token_embeddings/output_embeddings mismatch is unrelated to RTDetrLoss infeasible cost-matrix crash." + "reason": "Both touch Qwen3.5, but one is CPU offload rotary embedding and the other is a missing kernel decorator." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "DeepSpeed ZeRO3 deque pop error and Sam3LiteText auto_mappings CI failure are different bugs." + "reason": "DeepSpeed ZeRO3 deque crash vs stale auto_mappings config references; distinct bugs." }, { - "left": "issue:45003", - "right": "issue:45561", + "left": "issue:43295", + "right": "issue:45292", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is not the same as the pytest-xdist captured_info.txt race." + "reason": "Processor/tokenizer regression vs embedding resize propagation; different components and symptoms." }, { - "left": "issue:43295", - "right": "issue:45292", + "left": "issue:45003", + "right": "issue:45561", "accept": false, - "reason": "A processor/tokenizer regression in v4.57.5 is unrelated to a generic resize_token_embeddings/output_embeddings issue." + "reason": "sys.modules access bug vs pytest-xdist file race; unrelated issues." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "Qwen2.5-VL import failure and DeepSpeed/BertModel zero3 IndexError are unrelated." + "reason": "Missing Qwen2_5_VL import vs ZeRO3/BertModel IndexError; no shared root cause." }, { "left": "issue:43475", "right": "issue:46369", "accept": false, - "reason": "SAM 3 Video missing attribute and SAM3-LiteText memory overflow are different SAM-family problems with different symptoms and fixes." + "reason": "SAM3 Video missing attribute vs SAM3 LiteText memory overflow; different failures." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Both involve tied/resize embedding behavior, but one is about serialization differences by device and the other is specifically t5gemma decoder.embed_tokens not updating." + "reason": "Both involve tied/embedded weights, but one is serialization and the other is resize propagation in T5Gemma." }, { "left": "issue:44623", "right": "issue:46726", "accept": false, - "reason": "Missing files from processor.save_pretrained and AutoProcessor's unrecognized image processor error are separate save/load issues." + "reason": "save_pretrained missing files vs backend-missing AutoProcessor warning; distinct processor problems." }, { "left": "issue:44038", "right": "issue:46506", "accept": false, - "reason": "Qwen3-VL-Moe compatibility with transformers 5.0 and PPDocLayoutV3 CPU/CUDA result differences are unrelated." + "reason": "Qwen3-VL-Moe loading regression vs CPU/CUDA result mismatch in PPDocLayoutV3." }, { - "left": "issue:44242", - "right": "issue:44857", + "left": "issue:43526", + "right": "issue:46897", "accept": false, - "reason": "Load-balancing loss omission in a router-logits setting and an AMP/CUDA crash in LwDetrImageLoss are not the same bug." + "reason": "Different label-handling bugs in different models; not the same underlying change." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Strict config blocking granite_speech and a missing deepstack_visual_indexes field in Qwen3_5MoeVisionConfig are different config-validation problems." + "reason": "Both are strict-config complaints, but one blocks granite_speech loading while the other drops a Qwen field." }, { - "left": "issue:43526", - "right": "issue:46897", + "left": "issue:44242", + "right": "issue:44857", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to the Florence2 training-loss double-shift bug." + "reason": "Router/load-balancing loss gating vs LwDetr float16 AMP crash; unrelated." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Loading physical-intelligence/fast on transformers>=5.1.0 and hiding tokenizer decode errors when protobuf is absent are different failures." + "reason": "Physical-intelligence loading failure and protobuf tokenizer error masking are separate issues." }, { "left": "issue:29870", "right": "issue:30333", "accept": false, - "reason": "Bart generate() UnboundLocalError and MLflow not being reported after failed training are unrelated." + "reason": "Bart generate UnboundLocalError vs MLflow job reporting failure; unrelated." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "NemotronHForCausalLM Flash Attention 2.0 unsupported and the generic Flash Attention 2.0 issue are not the same concrete bug." + "reason": "Generic flash-attention topic vs a specific Nemotron FA2 support error; insufficient evidence of same bug." }, { "left": "issue:43526", "right": "issue:46894", "accept": false, - "reason": "BeitImageProcessorFast label reduction and CohereASR training-loss double-shift are different model-specific issues." + "reason": "Reduce-labels truncation vs ASR training-loss double-shift; similar pattern only at a very high level." }, { "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Qwen video-input regression and Windows/ROCm torch.distributed import/type errors are unrelated." + "reason": "Video input regression in Qwen models vs Windows ROCm/torch.distributed import issue; distinct code paths." }, { "left": "issue:42199", "right": "issue:46851", "accept": false, - "reason": "DETR-derived cardinality error without explicit background class is unrelated to Qwen3.5 variable-length position_ids handling." + "reason": "DETR cardinality error vs Qwen3.5 variable-length position_ids bug; unrelated." }, { - "left": "issue:44060", - "right": "issue:44077", + "left": "issue:45397", + "right": "issue:45656", "accept": false, - "reason": "Qwen3-Next tied-weights warning and patchtsmixer post_init restriction are different configuration/validation issues." + "reason": "Gemma-4 Zero3 from_pretrained problem vs optimizer stepping twice under DeepSpeed; different runtime bugs." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are not the same bug." + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 breakage; unrelated." }, { - "left": "issue:45397", - "right": "issue:45656", + "left": "issue:44368", + "right": "issue:44977", "accept": false, - "reason": "gemma_speech zero3 from_pretrained and optimizer.step being called twice with deepspeed are different deepspeed-related problems." + "reason": "Tie-word-embeddings warning during LoRA fine-tuning vs flash-attention generation failure; different issues." }, { - "left": "issue:44368", - "right": "issue:44977", + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Qwen3-Next tied-weights warning vs patchtsmixer post_init policy change; unrelated." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 pixel decoder dummy Conv2d vs sam3_video/sam3_tracker config mismatch; not the same bug." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict-config loading issue vs Windows ROCm torch.distributed dependency problem; unrelated." + }, + { + "left": "issue:43475", + "right": "issue:43540", "accept": false, - "reason": "Qwen3.5 tie_word_embeddings warning during LoRA fine-tuning and flash-attention generation failure are separate issues." + "reason": "SAM3 Video missing attribute vs Qwen3OmniMoe video input ValueError; different models and failures." } ] }, @@ -14015,9 +14012,6 @@ "nodes": [ "issue:42886", "issue:43065", - "issue:43408", - "issue:43475", - "issue:43540", "issue:43646", "issue:43701", "issue:43723", @@ -14033,14 +14027,15 @@ "issue:44623", "issue:44749", "issue:44871", - "issue:44877", "issue:44964", "issue:44995", "issue:45103", "issue:45137", "issue:45200", "issue:45229", + "issue:45375", "issue:45412", + "issue:45478", "issue:45563", "issue:45588", "issue:45593", @@ -14056,26 +14051,25 @@ "issue:46431", "issue:46710", "issue:46726", + "issue:46752", "issue:46832", "issue:46858", + "issue:47020", "issue:47068", "issue:47145" ], "soft_pairs": [ - "issue:43065|issue:43408", - "issue:44877|issue:46170", - "issue:43475|issue:43540", "issue:45412|issue:46123", - "issue:46170|issue:46832", "issue:42886|issue:45229", + "issue:46170|issue:46832", "issue:43723|issue:46710", "issue:43931|issue:44368", "issue:44188|issue:46202", "issue:44995|issue:45200", "issue:44368|issue:45563", "issue:46132|issue:47145", - "issue:44188|issue:46421", "issue:44483|issue:46202", + "issue:44188|issue:46421", "issue:43906|issue:46202", "issue:45593|issue:47068", "issue:44995|issue:45588", @@ -14091,216 +14085,218 @@ "issue:43646|issue:44162", "issue:44623|issue:45656", "issue:44964|issue:45137", - "issue:44871|issue:46132" + "issue:44871|issue:46132", + "issue:46752|issue:47020", + "issue:45375|issue:45478" ], - "prepared_review_unit_hash": "07b2f257e6dc0821db0c4fc86531dfb878ec7177c1919d3de89906ba8887a478", + "prepared_review_unit_hash": "44150872da82432727b80c397b87a6d5eff14f09e363b1b611675f60148d68d0", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36308, - "estimated_input_tokens": 9077, - "estimated_eval_tokens": 18410 + "soft_pair_count": 29, + "serialized_chars": 36046, + "estimated_input_tokens": 9012, + "estimated_eval_tokens": 18280 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is very heterogeneous: it contains a loose tokenizer/loading subgroup, but most items are clearly unrelated model, training, compilation, or pipeline bugs. None of the soft pairs look like true duplicates.", - "confidence": 0.31, - "canonical_issue_reason": "issue:43723 is the broadest and most central tokenizer-loading regression in the set, and it best anchors the small related subset around cache/save/load behavior.", + "summary": "This cluster is mostly heterogeneous: the issues span tokenizer loading, model/config regressions, training/optimizer behavior, cache semantics, and unrelated subsystem bugs. Only one soft pair looks like a true duplicate/near-duplicate: the StaticCache.get_seq_length contract bug and its _prefill slice-arithmetic failure are the same underlying issue.", + "confidence": 0.73, + "canonical_issue_reason": "issue:45987 is the clearest root-cause report: it names the API contract violation in StaticCache.get_seq_length, which directly explains the downstream failure described in issue:46858.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43723 is the most representative user-facing bug report among the related items; it is broad enough to serve as the anchor even though the overall cluster is noisy.", + "best_issue_reason": "issue:45987 is the best anchor for the only convincing duplicate pair because it states the underlying bug more generally, while issue:46858 is a specific manifestation.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43065", - "right": "issue:43408", - "accept": false, - "reason": "Different SAM3 bugs: dummy Conv2d in the pixel decoder vs model-type mismatch when instantiating sam3_tracker." - }, - { - "left": "issue:44877", - "right": "issue:46170", - "accept": false, - "reason": "Unrelated failures: strict config loading for granite_speech vs Windows ROCm import/type errors tied to torch.distributed." - }, - { - "left": "issue:43475", - "right": "issue:43540", - "accept": false, - "reason": "Both are video-related, but the concrete bugs are different code paths: missing fpn_position_embeddings in SAM3 Video vs Qwen3OmniMoe video-input ValueError." - }, { "left": "issue:45412", "right": "issue:46123", "accept": false, - "reason": "Different subsystems and symptoms: RT-DETR memory not released vs MaskGenerationPipeline dropping final partial-batch results." + "reason": "Different subsystems and failures: RT-DETR memory retention vs MaskGenerationPipeline partial-batch result dropping." }, { - "left": "issue:46170", - "right": "issue:46832", + "left": "issue:42886", + "right": "issue:45229", "accept": false, - "reason": "Windows/ROCm import dependency issue vs RT-DETR forward shape/feature-level mismatch; no shared root cause." + "reason": "Tokenizer offline cache loading is unrelated to Gemma4 multi-GPU CUDA OOM." }, { - "left": "issue:42886", - "right": "issue:45229", + "left": "issue:46170", + "right": "issue:46832", "accept": false, - "reason": "Tokenizer offline cache loading regression vs Gemma4 multi-GPU CUDA OOM; same broad area only at the highest level." + "reason": "Windows/ROCm import dependency issue is unrelated to an RTDetrModel forward shape mismatch." }, { "left": "issue:43723", "right": "issue:46710", "accept": false, - "reason": "Tokenizer loading failure in v5 vs accuracy regression after upgrading Transformers on DeepSeek; not the same bug." + "reason": "Both are version-related regressions, but one is tokenizer loading and the other is accuracy regression; not the same bug." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch vs a tie_word_embeddings warning; different severity and failure mode." + "reason": "Qwen3-VL weight-shape mismatch is unrelated to a warning about tie_word_embeddings in Qwen3.5 LoRA fine-tuning." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "torch.compile attention-kernel divergence vs Gemma pipeline roundtrip issue on \u2581; unrelated code paths." + "reason": "Different concrete problems: torch.compile attention-kernel divergence vs Gemma pipeline roundtrip failure." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "GlmMoeDsa stale indexer cache on second forward vs Gemma 4 mm_token_type_ids defaulting problem." + "reason": "GlmMoeDsa stale cache crash is unrelated to Gemma 4 mm_token_type_ids defaults." }, { "left": "issue:44368", "right": "issue:45563", "accept": false, - "reason": "Both mention warnings, but one is an embedding-tying config warning and the other is a stale generate() warning for num_return_sequences." + "reason": "A configuration warning during fine-tuning is unrelated to a stale generate warning for num_return_sequences." }, { "left": "issue:46132", "right": "issue:47145", "accept": false, - "reason": "AttentionInterface.register behavior change vs FP8 compressed-tensors loading without weight_scale; different mechanisms." + "reason": "AttentionInterface.register behavior change is unrelated to FP8 checkpoint loading missing weight_scale application." }, { - "left": "issue:44188", - "right": "issue:46421", + "left": "issue:44483", + "right": "issue:46202", "accept": false, - "reason": "torch.compile attention-kernel branching bug vs cache lazy initialization causing recompilation; related terminology, different root causes." + "reason": "Chat-completions API request rejection is unrelated to Gemma pipeline token roundtrip failure." }, { - "left": "issue:44483", - "right": "issue:46202", + "left": "issue:44188", + "right": "issue:46421", "accept": false, - "reason": "API request rejection in /v1/chat/completions vs Gemma pipeline roundtrip failure; not duplicates." + "reason": "Both mention compile/cache behavior, but one is attention-kernel branching and the other is cache lazy-init recompilation; not the same defect." }, { "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "Isolated reproduction of another issue vs a separate Gemma pipeline tokenization/roundtrip bug." + "reason": "An isolated reproduction of another issue is too generic to match the specific Gemma '\u2581' roundtrip failure." }, { "left": "issue:45593", "right": "issue:47068", "accept": false, - "reason": "D-FINE auxiliary-loss behavior when denoising is off vs a deepspeed host-synchronization concern in loss computation." + "reason": "D-FINE auxiliary-loss behavior and a deepspeed host-synchronization concern are different training-path bugs." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "Stale cache indexer on second forward vs flash_attention crash when s_aux is None; different layers." + "reason": "Second-forward-pass cache corruption in GlmMoeDsa is unrelated to flash_attention crashing on s_aux=None." }, { "left": "issue:45987", "right": "issue:46082", "accept": false, - "reason": "StaticCache sequence-length return-type bug vs LlamaConfig head_dim validation; unrelated." + "reason": "Cache sequence-length type violation is unrelated to LlamaConfig rejecting an explicit head_dim." }, { "left": "issue:45685", "right": "issue:46431", "accept": false, - "reason": "MPS histogram not implemented for Int in moe vs missing chat template error in processor loading." + "reason": "MPS histogram support failure is unrelated to missing chat-template errors in processors." }, { "left": "issue:44361", "right": "issue:44749", "accept": false, - "reason": "MLukeTokenizer AttributeError on tasks vs slowdown after upgrading Transformers; not the same defect." + "reason": "Tokenizer AttributeError on tasks is unrelated to a training throughput regression after upgrading Transformers." }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Both concern save_pretrained, but tokenizer_config class mismatch and processor missing files are separate serialization bugs." + "reason": "Both are save_pretrained-related, but one is tokenizer class metadata mismatch and the other is missing processor files; not the same concrete bug." }, { "left": "issue:45103", "right": "issue:46858", "accept": false, - "reason": "auto_docstring crash with future annotations vs static-cache seq-length returning a device tensor; different utilities and failures." + "reason": "auto_docstring annotation handling is unrelated to cache.get_seq_length returning a device tensor in prefill." }, { "left": "issue:43065", "right": "issue:46123", "accept": false, - "reason": "SAM3 pixel decoder dummy layer vs MaskGenerationPipeline final-batch drop; unrelated." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to the last-batch handling bug in MaskGenerationPipeline." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "resume_from_checkpoint key mismatch vs training slowdown after upgrade; no shared root cause." + "reason": "Checkpoint resume key mismatch is unrelated to a performance regression after version upgrade." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "Performance regression after upgrade vs optimizer.step being called twice under deepspeed; different bugs." + "reason": "Throughput slowdown and double optimizer.step under deepspeed are different training issues." }, { "left": "issue:43994", "right": "issue:46726", "accept": false, - "reason": "SigLIP2 nonsensical outputs with AutoModel/pipeline vs AutoProcessor backend-missing error; separate loading paths." + "reason": "SigLIP2 incorrect outputs and missing vision backend handling are different vision-stack problems." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "Transformers 5 custom model init breakage vs ESM2 being broken; too broad and not the same code path." + "reason": "Custom model initialization breakage in v5 and ESM2 being broken are not the same code-path issue." }, { "left": "issue:44623", "right": "issue:45656", "accept": false, - "reason": "Missing files from processor.save_pretrained vs deepspeed causing duplicate optimizer steps; unrelated." + "reason": "processor.save_pretrained file emission and deepspeed optimizer-step duplication are unrelated." }, { "left": "issue:44964", "right": "issue:45137", "accept": false, - "reason": "Phi-4 multimodal loading failure vs DeepSpeed ZeRO3 deque underflow; no overlap." + "reason": "Phi-4 multimodal loading failure is unrelated to DeepSpeed ZeRO3 deque underflow." }, { "left": "issue:44871", "right": "issue:46132", "accept": false, - "reason": "Gemma eos_token_id config inconsistency vs AttentionInterface.register behavior change; distinct model/config issues." + "reason": "Gemma-3 eos_token_id mismatch is unrelated to AttentionInterface.register behavior changes." + }, + { + "left": "issue:46752", + "right": "issue:47020", + "accept": false, + "reason": "Empty-list chat-template IndexError and XLMRobertaTokenizer/Unigram constructor mismatch are different tokenizer bugs." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 MoE, but one is a missing config field dropped by strict validation and the other is a broader from_pretrained error; insufficiently specific to merge." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": true, + "reason": "Same underlying cache API bug: get_seq_length returns a tensor instead of an int, and 46858 is the concrete _prefill failure this causes on non-CUDA backends." } ] }, @@ -14343,9 +14339,11 @@ "issue:43873", "issue:44038", "issue:44188", + "issue:44222", "issue:44351", "issue:44373", "issue:44623", + "issue:44661", "issue:44857", "issue:44869", "issue:44933", @@ -14357,28 +14355,24 @@ "issue:45356", "issue:45375", "issue:45397", - "issue:45478", "issue:45561", + "issue:45593", "issue:45753", "issue:46326", "issue:46399", "issue:46463", "issue:46726", - "issue:46752", "issue:46821", "issue:46851", "issue:46877", "issue:47000", - "issue:47020", "issue:47059", "issue:47217" ], "soft_pairs": [ - "issue:46752|issue:47020", - "issue:45375|issue:45478", - "issue:42994|issue:43482", - "issue:45561|issue:47217", "issue:43475|issue:46821", + "issue:45561|issue:47217", + "issue:42994|issue:43482", "issue:43475|issue:43643", "issue:44188|issue:46851", "issue:45230|issue:45397", @@ -14388,136 +14382,133 @@ "issue:28282|issue:30990", "issue:44038|issue:44623", "issue:39692|issue:46726", - "issue:46877|issue:47000", "issue:41628|issue:44933", - "issue:42994|issue:43525", + "issue:46877|issue:47000", "issue:46463|issue:46877", + "issue:42994|issue:43525", "issue:44869|issue:45356", "issue:43723|issue:46326", "issue:42371|issue:45092", - "issue:43784|issue:46399", "issue:43873|issue:44945", - "issue:41628|issue:44351", + "issue:43784|issue:46399", "issue:45137|issue:47000", + "issue:41628|issue:44351", "issue:36010|issue:41628", "issue:42994|issue:43824", "issue:44038|issue:45310", "issue:44945|issue:45753", + "issue:45561|issue:47059", "issue:44857|issue:46877", - "issue:45561|issue:47059" + "issue:44222|issue:45310", + "issue:44661|issue:47059", + "issue:45593|issue:46877" ], - "prepared_review_unit_hash": "a647a83da25779aa027540ab2a35d1b2365361239e2d3ff24b047d9958abb973", + "prepared_review_unit_hash": "5d5c2d822108bac561f4b58ad5a7a5a04e834ca91f3bf923d864288105b48853", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36596, - "estimated_input_tokens": 9149, - "estimated_eval_tokens": 18554 + "soft_pair_count": 32, + "serialized_chars": 36847, + "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 mostly a loose collection of Transformers v5/import/API regression reports and model-specific bugs. The soft pairs look thematically similar in a few cases, but they do not appear to be the same concrete underlying defect, so I would not merge any of them.", - "confidence": 0.35, - "canonical_issue_reason": "issue:41628 is the most broadly representative concrete regression in the set: a core import failure in the public Transformers API, which matches the cluster\u2019s repeated import/compatibility theme better than the very model-specific reports.", + "summary": "This cluster is mostly heterogeneous: many unrelated import, generation, quantization, vision, and testing regressions got linked only by broad Transformers v5 similarity. I found one plausible duplicate-style pair in the Qwen3.5 MoE loading/config path, but the rest should stay separate.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45310 is the broadest user-facing report in the only clearly related subthread: a Qwen3.5 MoE from_pretrained regression on transformers>=5.4.0. It is a better umbrella than the more specific config-field report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:41628 is the best single issue to anchor the cluster because it is specific enough to be actionable yet general enough to represent the recurring missing-export/import-breakage pattern across nearby reports.", + "best_issue_reason": "45310 is the best global issue anchor because it names the affected model family and the concrete loading regression, and it can subsume the more specific missing-field symptom in 45375.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:46752", - "right": "issue:47020", - "accept": false, - "reason": "Empty-list chat-template crash and XLMRobertaTokenizer/Unigram init bug are different code paths and failure modes." - }, - { - "left": "issue:45375", - "right": "issue:45478", - "accept": false, - "reason": "Both mention Qwen3.5/MoE, but one is a missing config field under @strict and the other is a generic from_pretrained failure; not enough to treat as the same bug." - }, - { - "left": "issue:42994", - "right": "issue:43482", + "left": "issue:43475", + "right": "issue:46821", "accept": false, - "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are opposite directions and likely distinct issues." + "reason": "Both are AttributeError reports, but they hit different models, different attributes, and different code paths." }, { "left": "issue:45561", "right": "issue:47217", "accept": false, - "reason": "pytest-xdist file race in testing utils is unrelated to VL image-placeholder handling." + "reason": "One is a pytest-xdist/test utility race; the other is a VL prompt/image-placeholder inference bug." }, { - "left": "issue:43475", - "right": "issue:46821", + "left": "issue:42994", + "right": "issue:43482", "accept": false, - "reason": "Different model families and unrelated missing-attribute crashes; only a superficial AttributeError similarity." + "reason": "Both involve quantization-related workflows, but one is saving and the other is loading a different model format; not the same bug." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "Sam3 vision encoder output missing a field is not the same as trust_remote_code dropping fields from AutoConfig results." + "reason": "SAM3 encoder attribute failure and AutoConfig missing fields are unrelated symptoms with different components." }, { "left": "issue:44188", "right": "issue:46851", "accept": false, - "reason": "torch.compile attention-kernel divergence and Qwen3.5 variable-length position_ids behavior are different subsystems." + "reason": "Different subsystems and failure modes: torch.compile attention-kernel divergence vs Qwen3.5 variable-length position_ids handling." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "The first is an empty generic report; the second is a specific gemma-4 from_pretrained/zero3 issue." + "reason": "Too generic to merge: one is a placeholder bug report, the other is a specific gemma-4 ZeRO3 from_pretrained issue." }, { "left": "issue:44373", "right": "issue:46851", "accept": false, - "reason": "A wrong docstring for position_ids is documentation-only, while the other is a runtime bug in Qwen3.5 variable-length inputs." + "reason": "A wrong docstring for position_ids is not the same as a runtime failure in variable-length Qwen3.5 inputs." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Both involve generation internals, but one is sync_gpus bookkeeping and the other is chunked_prefill cache handling; not the same concrete defect." + "reason": "Both mention generation internals, but they describe different failures in different generation paths." }, { "left": "issue:45310", "right": "issue:45375", - "accept": false, - "reason": "Both are Qwen3.5/MoE-related, but one is a broad from_pretrained error and the other is a specific missing config field; too little evidence they are the same bug." + "accept": true, + "reason": "Same underlying Qwen3.5 MoE loading/config regression: 45310 reports the v5.4.0 from_pretrained failure, while 45375 identifies the missing deepstack_visual_indexes field dropped by strict config handling." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "PyTorch-missing ImportError is unrelated to Sentence Transformers loading hangs." + "reason": "PyTorch-missing import error and Sentence Transformers hanging on load are different problems." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL-Moe bug and processor.save_pretrained file omission are different parts of the stack." + "reason": "Qwen3-VL-Moe bug and processor.save_pretrained missing files are separate issues." }, { "left": "issue:39692", "right": "issue:46726", "accept": false, - "reason": "SigLIP2 doc-example errors and AutoProcessor missing-backend behavior are both vision-related but not the same failure." + "reason": "Both touch vision/processor docs, but one is a broken example and the other is an AutoProcessor backend-missing error." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they concern different symbols and likely different export paths." }, { "left": "issue:46877", @@ -14526,108 +14517,123 @@ "reason": "Import-time docstring lint output and RTDetrLoss infeasible cost matrix are unrelated." }, { - "left": "issue:41628", - "right": "issue:44933", + "left": "issue:46463", + "right": "issue:46877", "accept": false, - "reason": "Both are import-related, but they involve different missing symbols and different modules; not clearly the same underlying regression." + "reason": "Incorrect T5 validation messaging is not the same bug as doc-lint output at import time." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized model saving and Llama4Config missing pad_token_id are unrelated." - }, - { - "left": "issue:46463", - "right": "issue:46877", - "accept": false, - "reason": "T5 decoder-input validation messaging and auto_docstring stdout noise are unrelated." + "reason": "Quantized save failure and Llama4Config pad_token_id absence are different model/config bugs." }, { "left": "issue:44869", "right": "issue:45356", "accept": false, - "reason": "Whisper trailing-replacement crash and Kimi-K2.5 tokenizer regression are distinct tokenizer bugs." + "reason": "Both involve tokenization/decoding, but they affect different models and different decoding failures." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "AutoTokenizer loading failure and mistral_common continue_final_message bug affect different APIs and code paths." + "reason": "Tokenizer loading in v5 and mistral_common continue_final_message handling are separate code paths." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 settings API guidance and old InternVL2 remote-code/meta-init incompatibility are unrelated." - }, - { - "left": "issue:43784", - "right": "issue:46399", - "accept": false, - "reason": "Sentence-transformers import NameError and Qwen3.5 rotary-kernel registration failure are different errors." + "reason": "TF32 settings guidance and InternVL2 meta-initialization incompatibility are unrelated." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization offloading behavior and incorrect pipeline-parallelism output are not the same bug." + "reason": "Quantization offloading and pipeline-parallel LLM output divergence are different runtime problems." }, { - "left": "issue:41628", - "right": "issue:44351", + "left": "issue:43784", + "right": "issue:46399", "accept": false, - "reason": "Both are missing-import reports, but AutoImageProcessor and HybridCache are different exports and likely different breakages." + "reason": "sentence-transformers import NameError and Qwen3.5 kernel registration failure are unrelated." }, { "left": "issue:45137", "right": "issue:47000", "accept": false, - "reason": "ZeRO3 deque underflow and RTDetrLoss cost-matrix infeasibility are unrelated runtime errors." + "reason": "DeepSpeed ZeRO3 deque underflow and RTDetrLoss infeasible cost matrix are different failures." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are top-level import errors, but for different symbols and likely different compatibility fixes." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but one is GenerationMixin and the other is AutoImageProcessor; too broad to merge." + "reason": "GenerationMixin import path breakage and AutoImageProcessor import failure are separate API regressions." }, { "left": "issue:42994", "right": "issue:43824", "accept": false, - "reason": "Quantized saving failure and Qwen2_5_VLForConditionalGeneration import failure are different classes of problems." + "reason": "Quantized save failure and missing Qwen2.5-VL class import are unrelated." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Qwen3-VL-Moe bug and Qwen3.5-MoE from_pretrained error are related only at a very broad model-family level." + "reason": "Different Qwen model variants and different failure points; too broad to treat as one bug." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallelism output corruption and Qwen3_5 image infinite loop are distinct symptoms and paths." + "reason": "Pipeline-parallel output errors and Qwen3.5 image-specific looping are not the same code-path issue." + }, + { + "left": "issue:45561", + "right": "issue:47059", + "accept": false, + "reason": "A test race on captured_info.txt is unrelated to a huggingface_hub CLI compatibility break." }, { "left": "issue:44857", "right": "issue:46877", "accept": false, - "reason": "AMP/CUDA loss crash and import-time doc-lint stdout are unrelated." + "reason": "Float16 AMP CUDA crash in LwDetrLoss is unrelated to import-time docstring diagnostics." }, { - "left": "issue:45561", + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 save_pretrained for MoE and Qwen3.5 from_pretrained loading failure are different operations and likely different fixes." + }, + { + "left": "issue:44661", "right": "issue:47059", "accept": false, - "reason": "pytest-xdist race in tests and HFCliTyperGroup attribute breakage are unrelated." + "reason": "TOKENIZER_MAPPING_NAMES model-generation failure and CLI completion regression are unrelated." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "D-FINE auxiliary-loss logic and doc-lint stdout noise are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, - "error_kind": null, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats the cluster as mostly heterogeneous and only elevates the 45310/45375 Qwen3.5 MoE pair as a plausible duplicate-style match. The other soft-edge rejections are cautious and supported by the issue titles/descriptions rather than overstated evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, "error_message": null }, { @@ -14656,6 +14662,8 @@ "issue:43479", "issue:43704", "issue:43723", + "issue:43784", + "issue:43824", "issue:43976", "issue:44016", "issue:44186", @@ -14663,8 +14671,8 @@ "issue:44496", "issue:44568", "issue:44623", - "issue:44661", "issue:44743", + "issue:44779", "issue:44792", "issue:44869", "issue:44933", @@ -14674,7 +14682,6 @@ "issue:45405", "issue:45406", "issue:45561", - "issue:45593", "issue:45600", "issue:45712", "issue:45812", @@ -14692,11 +14699,8 @@ "issue:47085" ], "soft_pairs": [ - "issue:44222|issue:45310", - "issue:45593|issue:46877", - "issue:44661|issue:47059", - "issue:45081|issue:45103", "issue:45406|issue:46202", + "issue:45081|issue:45103", "issue:46202|issue:46701", "issue:43232|issue:44792", "issue:42673|issue:43704", @@ -14708,9 +14712,9 @@ "issue:30333|issue:33290", "issue:43976|issue:44933", "issue:30990|issue:35141", - "issue:43479|issue:46829", "issue:42898|issue:44568", "issue:42175|issue:45712", + "issue:43479|issue:46829", "issue:44568|issue:47085", "issue:44623|issue:45310", "issue:42673|issue:46693", @@ -14721,222 +14725,218 @@ "issue:43723|issue:44568", "issue:43479|issue:44869", "issue:44016|issue:46231", - "issue:39692|issue:46909" + "issue:39692|issue:46909", + "issue:43784|issue:43824", + "issue:42898|issue:44779" ], - "prepared_review_unit_hash": "4b04e9dec9b31c12b5e041946a6349d498c765c1fbdcafd3d6914dbc7b58db0a", + "prepared_review_unit_hash": "23dcb6de393894fe5ae26128f6873d120ae916b9062cf47874102e89307772ad", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36287, - "estimated_input_tokens": 9072, - "estimated_eval_tokens": 18400 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36605, + "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 unrelated issue noise; only one soft pair is a clear duplicate. The Qwen3ForCausalLM VRAM leak reports are the same bug, while the other pairs are at best thematically related but describe different code paths or symptoms.", - "confidence": 0.95, - "canonical_issue_reason": "issue:42673 is the broader/original report of the Qwen3ForCausalLM multi-dataloader-thread VRAM leak, and issue:43704 appears to be the same bug described with the same title and symptom.", + "summary": "Only one soft edge is a clear duplicate: the Qwen3ForCausalLM VRAM leak report appears twice with the same title. The other candidate pairs are mostly broad-theme matches (tokenizers, Gemma, FP8, docs, generation) but describe different concrete failures and should stay separate.", + "confidence": 0.96, + "canonical_issue_reason": "Use issue:42673 as canonical: it is the earlier, fuller report of the exact same Qwen3ForCausalLM multi-dataloader VRAM leak, and issue:43704 is a near-verbatim duplicate with less context.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42673 is the best issue to keep as canonical because it is the earlier, fuller report of the duplicate VRAM leak problem.", + "best_issue_reason": "issue:42673 is the best issue fit because it most clearly states the bug, has the richer discussion trail, and cleanly represents the duplicated Qwen3ForCausalLM VRAM leak problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44222", - "right": "issue:45310", - "accept": false, - "reason": "Both mention Qwen3.5/MoE/FP8, but one is about save_pretrained and the other about from_pretrained loading failure; not clearly the same defect." - }, - { - "left": "issue:45593", - "right": "issue:46877", - "accept": false, - "reason": "Different subsystems and symptoms: D-FINE auxiliary losses vs docstring diagnostics printing at import." - }, - { - "left": "issue:44661", - "right": "issue:47059", + "left": "issue:45406", + "right": "issue:46202", "accept": false, - "reason": "Unrelated: add-new-model-like/tokenizer mapping logic vs a CLI regression with huggingface_hub completion APIs." + "reason": "Both involve Gemma, but one is a processor init bug and the other is a pipeline roundtrip/tokenization failure; different code paths and symptoms." }, { "left": "issue:45081", "right": "issue:45103", "accept": false, - "reason": "Different bugs in different utilities: Mistral tokenizer regex patching vs auto_docstring future-annotations AttributeError." - }, - { - "left": "issue:45406", - "right": "issue:46202", - "accept": false, - "reason": "Both involve Gemma-related code, but one is a processor attribute crash in serve and the other is a tokenization roundtrip failure; different failures." + "reason": "Both are AttributeErrors, but one is a Mistral tokenizer regex patch failure and the other is auto-docstring handling of future annotations; unrelated bugs." }, { "left": "issue:46202", "right": "issue:46701", "accept": false, - "reason": "Different Gemma pipeline issues with different triggers and code paths; not enough evidence of one underlying bug." + "reason": "Gemma-related, but one is a tokenizer/pipeline roundtrip issue and the other is a DiffusionGemmaGenerationOutput pipeline failure; not the same defect." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Different models and paths: generation kwargs handling vs a janus image-generation test failure." + "reason": "Generation kwargs handling after sync_gpus is unrelated to the janus image generation test failure." }, { "left": "issue:42673", "right": "issue:43704", "accept": true, - "reason": "Exact same title and symptom about Qwen3ForCausalLM leaking VRAM in multiple dataloader threads; clear duplicate." + "reason": "Same title and same underlying symptom: Qwen3ForCausalLM leaks VRAM when used in multiple dataloader threads." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "Both mention FP8, but one is an accuracy issue and the other is save_pretrained behavior for MoE; different bugs." + "reason": "Both mention FP8, but one is an accuracy regression and the other is a save_pretrained/moe serialization issue." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Completely different issues: missing TensorFlow backend dependency vs a missing import in image_utils." + "reason": "TensorFlow backend packaging and a missing image_utils import are unrelated." }, { "left": "issue:45405", "right": "issue:46877", "accept": false, - "reason": "PEFT version bump incompatibility is unrelated to docstring stdout logging." + "reason": "A dependency version bump issue and an auto_docstring stdout logging issue are unrelated." }, { "left": "issue:44743", "right": "issue:47068", "accept": false, - "reason": "Different components and failure modes: Qwen3.5 recurrent-state reset vs a deepspeed host-synchronization concern." + "reason": "A Qwen3.5 recurrent-state caching bug and a deepspeed loss synchronization concern are different code paths." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Unrelated: model config loading error vs xdist race in captured_info.txt." + "reason": "Model loading/config recognition and a pytest-xdist file race are unrelated." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "Different training/runtime failures: MLflow reporting after failure vs OOM with Adafactor in deepspeed." + "reason": "MLflow reporting on failed training jobs is unrelated to a Sentence Transformers loading hang." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version support issue vs unrelated missing import bug." + "reason": "Python version support and a missing image_utils import are separate problems." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Different models and code paths: sentence-transformers loading hang vs embedding resize/post_init reinitialization." - }, - { - "left": "issue:43479", - "right": "issue:46829", - "accept": false, - "reason": "Different multimodal model bugs with distinct symptoms and outputs; no shared underlying defect." + "reason": "A loading hang and an embedding reinitialization bug are different failures." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer regressions in v5, but one is about cleanup behavior and the other about add_special_tokens/BOS-EOS handling; not the same bug." + "reason": "Both are tokenizer regressions in v5, but one concerns clean_up_tokenization_spaces behavior while the other is BOS/EOS insertion for a specific tokenizer." }, { "left": "issue:42175", "right": "issue:45712", "accept": false, - "reason": "Different packaging/tooling issues: TensorFlow backend omission vs leftover dummy classes leaking into dir/check_repo." + "reason": "Backend dependency packaging and leftover dummy classes are unrelated." + }, + { + "left": "issue:43479", + "right": "issue:46829", + "accept": false, + "reason": "Phi4Multimodal config initialization and VideoPrism output field wiring are different model-specific issues." }, { "left": "issue:44568", "right": "issue:47085", "accept": false, - "reason": "One is a specific tokenizer behavior regression, the other is a broader thread-safety regression; not the same concrete defect." + "reason": "A tokenizer special-tokens regression and a tokenizer thread-safety regression are different bugs." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Processor.save_pretrained missing files is unrelated to Qwen3.5 MoE from_pretrained loading failure." + "reason": "Processor save_pretrained missing files and Qwen3.5 MoE from_pretrained errors are unrelated." }, { "left": "issue:42673", "right": "issue:46693", "accept": false, - "reason": "Both are performance/memory related, but VRAM leak in dataloader threads is unrelated to FA2 long-context generation slowdown." + "reason": "Both are performance-related, but VRAM leakage in dataloader threads is not the same as FA2 long-context generation slowdown." }, { "left": "issue:44186", "right": "issue:47085", "accept": false, - "reason": "LayoutLMv2 NER/padding crashes are a specific tokenizer bug, not the same as a general thread-safety regression." + "reason": "LayoutLMv2Tokenizer NER/padding crashes are separate from fast tokenizer thread-safety regressions." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Different behaviors in different pipelines: attention-mask construction vs partial-batch result dropping." + "reason": "Gemma4 attention-mask construction and MaskGenerationPipeline final-batch handling are unrelated." }, { "left": "issue:45600", "right": "issue:47059", "accept": false, - "reason": "Broken auto_mappings references and CLI completion failures are unrelated CI issues." + "reason": "A broken auto_mappings reference and a huggingface_hub CLI compatibility failure are unrelated CI issues." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different behaviors/models: clean_up_tokenization_spaces vs Granite token IDs." + "reason": "Both are tokenizer regressions, but the affected behavior and model scope differ; one is cleanup-space behavior, the other is wrong token IDs for Granite models." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Generic tokenizer-loading problem vs a specific mdeberta add_special_tokens regression; not clearly the same bug." + "reason": "Tokenizer loading failure in v5 and BOS/EOS special-token insertion failure are different symptoms and likely different fixes." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Different model families and different failure mechanisms: multimodal config defaults vs Whisper timestamp decoding." + "reason": "Phi4Multimodal config defaults and Whisper word-timestamp decoding are unrelated." }, { "left": "issue:44016", "right": "issue:46231", "accept": false, - "reason": "Notebook syntax error and a doc parameter-name mismatch are unrelated documentation issues." + "reason": "A notebook syntax error and a documentation parameter-name mismatch are not the same issue." }, { "left": "issue:39692", "right": "issue:46909", "accept": false, - "reason": "Both are doc-example problems, but for different models and distinct errors; not a duplicate." + "reason": "Both are broken documentation examples, but they target different models and failure modes." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Importing sentence-transformers fails with NameError, while the other is an ImportError for Qwen2.5-VL; unrelated imports." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but the reports describe different models and failure patterns rather than one shared concrete bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: issue:42673 and issue:43704 have the same title and clearly look like duplicates, and the canonical choice is reasonably supported by earlier creation time and longer body length. The other soft-edge rejections stay conservative and do not overclaim sameness from superficial similarity alone." + "feedback": "The analysis is grounded in the packet. The Qwen3ForCausalLM pair is a clear duplicate with identical titles, and choosing issue:42673 as canonical is well-supported by the earlier timestamp and richer discussion trail. The remaining soft-edge rejections stay conservative and do not overstate evidence." }, "evaluator_used": true, "retried": false, @@ -14963,8 +14963,10 @@ "issue:30333", "issue:34689", "issue:35141", + "issue:41628", "issue:42222", "issue:42673", + "issue:42757", "issue:42796", "issue:42831", "issue:42898", @@ -14973,8 +14975,6 @@ "issue:43479", "issue:43644", "issue:43723", - "issue:43784", - "issue:43824", "issue:43874", "issue:44060", "issue:44075", @@ -14989,7 +14989,6 @@ "issue:44485", "issue:44623", "issue:44734", - "issue:44779", "issue:44843", "issue:44933", "issue:45092", @@ -14997,6 +14996,7 @@ "issue:45125", "issue:45200", "issue:45245", + "issue:45310", "issue:45356", "issue:45405", "issue:45656", @@ -15007,10 +15007,8 @@ "issue:47085" ], "soft_pairs": [ - "issue:43784|issue:43824", - "issue:42898|issue:44779", - "issue:43479|issue:44079", "issue:27957|issue:30333", + "issue:43479|issue:44079", "issue:42673|issue:47085", "issue:44843|issue:47085", "issue:30333|issue:34689", @@ -15018,8 +15016,8 @@ "issue:42898|issue:43295", "issue:44295|issue:45656", "issue:42222|issue:45200", - "issue:44734|issue:45103", "issue:42898|issue:45356", + "issue:44734|issue:45103", "issue:30333|issue:35141", "issue:22355|issue:28282", "issue:44222|issue:44297", @@ -15027,8 +15025,8 @@ "issue:43116|issue:44190", "issue:45405|issue:45712", "issue:44297|issue:45656", - "issue:44933|issue:45712", "issue:42831|issue:44623", + "issue:44933|issue:45712", "issue:44448|issue:46132", "issue:44222|issue:44623", "issue:44060|issue:45125", @@ -15036,216 +15034,218 @@ "issue:43479|issue:43644", "issue:45245|issue:46132", "issue:44373|issue:44485", - "issue:44360|issue:46762" + "issue:44360|issue:46762", + "issue:41628|issue:42757", + "issue:44297|issue:45310" ], - "prepared_review_unit_hash": "883a3681729d5b30d32132c8dd76cd348f7cb51ea22809d1334033b529c780ce", + "prepared_review_unit_hash": "2a8b63805be1d7732f9c7b68ed645c6a547d784c07b9d6119a8f525464433380", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36665, - "estimated_input_tokens": 9167, - "estimated_eval_tokens": 18590 + "serialized_chars": 36575, + "estimated_input_tokens": 9144, + "estimated_eval_tokens": 18544 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are a loose collection of unrelated issues rather than one duplicate cluster. All soft duplicate candidates are false positives; they share at most broad themes like \u201ctokenizer\u201d or \u201cv5 regression\u201d but not the same concrete bug or fix path.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "No soft pair looks like the same underlying bug or change; the set is a heterogeneous mix of tokenizer/import, multimodal, optimizer, serving, and training issues. If a representative issue is needed, issue 44843 is the strongest single example because it is a concrete, user-facing tokenizer loading regression with clear reproduction and broader impact.", + "confidence": 0.92, + "canonical_issue_reason": "issue:44843 is the best representative issue here: it describes a concrete core regression in AutoTokenizer loading, has clear failure mode (HF_HUB_OFFLINE), and is more broadly relevant than the more niche or model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:44843 is the most suitable single issue to anchor the cluster, but only as a representative example\u2014not as a true duplicate of the rest.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43784", - "right": "issue:43824", - "accept": false, - "reason": "Both are import failures, but for different symbols and different failure modes; not the same underlying bug." - }, - { - "left": "issue:42898", - "right": "issue:44779", + "left": "issue:27957", + "right": "issue:30333", "accept": false, - "reason": "Both involve tokenizer behavior in v5, but one is a global cleanup-spaces change and the other is a DeepSeek-specific tokenizer regression." + "reason": "Different subsystems and failures: Flash Attention 2 model support vs MLflow reporting on failed training jobs." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Different code paths: multimodal config initialization vs ModelOutput key assignment." - }, - { - "left": "issue:27957", - "right": "issue:30333", - "accept": false, - "reason": "Unrelated subsystems: Flash Attention model support vs MLflow reporting on failed training jobs." + "reason": "Unrelated bugs: multimodal config initialization vs ModelOutput key assignment semantics." }, { "left": "issue:42673", "right": "issue:47085", "accept": false, - "reason": "Threading is mentioned in both, but one is a VRAM leak in Qwen3ForCausalLM dataloader threads and the other is tokenizer thread safety." + "reason": "Both mention threading, but one is a Qwen3 VRAM leak in dataloader threads and the other is fast tokenizer thread safety; different code paths and bugs." }, { "left": "issue:44843", "right": "issue:47085", "accept": false, - "reason": "Offline hub lookup in AutoTokenizer patching is unrelated to Rust tokenizer thread safety." + "reason": "Tokenizer offline-mode regression is not the same as Rust tokenizer thread-safety regression." }, { "left": "issue:30333", "right": "issue:34689", "accept": false, - "reason": "MLflow job-status reporting and Llama 3.2 Vision model loading are different bugs." + "reason": "MLflow job-status reporting and Llama 3.2 Vision model loading are unrelated issues." }, { "left": "issue:43874", "right": "issue:45092", "accept": false, - "reason": "Different model families and failures: missing image-processor method vs remote-code meta-init incompatibility." + "reason": "Both are multimodal/model-loading bugs, but one is a missing image processor method and the other is a meta-initialization incompatibility; not the same underlying defect." }, { "left": "issue:42898", "right": "issue:43295", "accept": false, - "reason": "Both are v5 regressions, but one is token cleanup behavior and the other is processor/tokenizer API handling for images." + "reason": "Both are v5 regressions around token/processor behavior, but they affect different APIs and code paths." }, { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "Position_ids buffer access and optimizer stepping with deepspeed are unrelated." + "reason": "Position_ids buffer access error and deepspeed optimizer double-step are unrelated." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "Different models and symptoms: vitpose breakage vs Gemma 4 multimodal token defaults." + "reason": "Different model families and different multimodal setup bugs; not the same change." }, { - "left": "issue:44734", - "right": "issue:45103", + "left": "issue:42898", + "right": "issue:45356", "accept": false, - "reason": "KV-cache indexing crash in serve is unrelated to auto_docstring crashing on future annotations." + "reason": "Both are tokenizer regressions, but one is clean_up_tokenization_spaces behavior and the other is Kimi-K2.5 codec/regex handling." }, { - "left": "issue:42898", - "right": "issue:45356", + "left": "issue:44734", + "right": "issue:45103", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and stem from different causes." + "reason": "Serving KV-cache indexing crash and auto-docstring annotation crash are unrelated." }, { "left": "issue:30333", "right": "issue:35141", "accept": false, - "reason": "MLflow failure reporting is unrelated to embedding reinitialization after resizing tokens." + "reason": "MLflow reporting vs token-embedding resize/post_init reinitialization are different bugs." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Different import errors: missing transformers.onnx module vs missing PyTorch backend." + "reason": "Both are import errors, but one is missing transformers.onnx and the other is missing PyTorch for AutoModel; unrelated causes." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "Both touch save_pretrained, but one is FP8 MoE serialization and the other is tokenizer config metadata." + "reason": "FP8 save_pretrained behavior and tokenizer_config class naming are unrelated." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "Optimizer argument handling and documentation parameter-name mismatch are different issues." + "reason": "Optimizer argument handling and docstring parameter mismatch are not the same bug." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Both are example-script failures, but for different scripts and different data-loading logic." + "reason": "Example-script failures in different scripts with different root causes; too broad to merge." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "Unrelated: unreleased PEFT version pin vs leftover dummy classes in a generated file." + "reason": "PEFT version pinning and leftover dummy classes are unrelated maintenance issues." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer save metadata mismatch is unrelated to deepspeed optimizer double-stepping." + "reason": "Tokenizer config serialization and deepspeed optimizer stepping are unrelated." }, { - "left": "issue:44933", - "right": "issue:45712", + "left": "issue:42831", + "right": "issue:44623", "accept": false, - "reason": "Missing image_utils import and leftover dummy classes are different maintenance problems." + "reason": "FineGrainedFP8 accuracy vs processor.save_pretrained missing files are unrelated." }, { - "left": "issue:42831", - "right": "issue:44623", + "left": "issue:44933", + "right": "issue:45712", "accept": false, - "reason": "FP8 accuracy and processor.save_pretrained missing files do not share a concrete fix path." + "reason": "Bad image_utils import and dummy class leakage are different export/import problems." }, { "left": "issue:44448", "right": "issue:46132", "accept": false, - "reason": "Pegasus v4/v5 output differences and AttentionInterface.register behavior changes are different regressions." + "reason": "Pegasus v4/v5 output difference and AttentionInterface.register behavior change are unrelated." }, { "left": "issue:44222", "right": "issue:44623", "accept": false, - "reason": "MoE FP8 save_pretrained bug is unrelated to processor file-saving omissions." + "reason": "Both involve saving, but one is FP8 moe save_pretrained and the other is processor save_pretrained file omission; different code paths." }, { "left": "issue:44060", "right": "issue:45125", "accept": false, - "reason": "Both are Qwen3-related, but one is a tied-weights warning and the other is missing tensor-parallel metadata." + "reason": "Both involve Qwen3 variants, but one is a tied-weights warning bug and the other is missing tensor-parallel plan support." }, { "left": "issue:42796", "right": "issue:43723", "accept": false, - "reason": "Model-specific BPE marker handling vs generic AutoTokenizer loading in v5 are not the same bug." + "reason": "Tokenizer BPE marker handling and AutoTokenizer.from_pretrained loading regression are different defects." }, { "left": "issue:43479", "right": "issue:43644", "accept": false, - "reason": "Default multimodal config initialization and non-persistent buffer junk fill are unrelated." + "reason": "Multimodal config defaults and non-persistent buffer corruption are unrelated." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "Category cardinality limits and AttentionInterface behavior changes are unrelated." + "reason": "Category-count limit error and AttentionInterface.register behavior change are unrelated." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "Docstring wording for position_ids is not the same as GLM-5 RoPE implementation." + "reason": "Docstring typo/accuracy issue is not the same as GLM-5 RoPE implementation." }, { "left": "issue:44360", "right": "issue:46762", "accept": false, - "reason": "DSA indexer activation gap and Minimax attention numerical drift are different code paths." + "reason": "Different algorithmic components with different correctness concerns; no shared code path." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Two distinct import errors from different packages and symbols." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer class serialization bug and Qwen3.5 MoE from_pretrained regression are different issues." } ] }, @@ -15274,7 +15274,6 @@ "issue:30990", "issue:38175", "issue:41628", - "issue:42757", "issue:42831", "issue:42907", "issue:43122", @@ -15291,17 +15290,18 @@ "issue:44263", "issue:44279", "issue:44295", - "issue:44297", "issue:44360", "issue:44373", "issue:44484", "issue:44623", + "issue:44829", "issue:44977", "issue:45071", "issue:45092", "issue:45310", "issue:45357", "issue:45834", + "issue:46002", "issue:46050", "issue:46082", "issue:46209", @@ -15318,12 +15318,10 @@ "issue:47059" ], "soft_pairs": [ - "issue:41628|issue:42757", - "issue:44297|issue:45310", "issue:45071|issue:46082", "issue:43723|issue:44977", - "issue:43504|issue:47059", "issue:44222|issue:44484", + "issue:43504|issue:47059", "issue:46614|issue:46932", "issue:43122|issue:46209", "issue:44263|issue:45092", @@ -15334,222 +15332,217 @@ "issue:46209|issue:46431", "issue:30990|issue:38175", "issue:44190|issue:46082", - "issue:43475|issue:46305", "issue:44263|issue:44360", + "issue:43475|issue:46305", "issue:46305|issue:46655", "issue:46369|issue:46909", "issue:43986|issue:46726", "issue:42831|issue:45310", "issue:22355|issue:27957", "issue:42907|issue:46050", - "issue:45092|issue:45357", "issue:44279|issue:46328", + "issue:45092|issue:45357", "issue:42831|issue:44038", "issue:44261|issue:45092", - "issue:41628|issue:43824" + "issue:41628|issue:43824", + "issue:44829|issue:46002" ], - "prepared_review_unit_hash": "6d0adb2b3cadd4647088841d6e5125ac821a51d7f4ce133d663de876b85689eb", + "prepared_review_unit_hash": "45f997856c53e49b91f74c802445312eabf4d63bb260daff951170415479b1dd", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35423, - "estimated_input_tokens": 8856, - "estimated_eval_tokens": 17968 + "soft_pair_count": 28, + "serialized_chars": 35169, + "estimated_input_tokens": 8793, + "estimated_eval_tokens": 17842 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly noisy title-similarity matches across unrelated bugs. A few items share a broad Transformers v5 loading/save-pretrained theme, but they are different failure modes and should stay separate; none of the soft pairs look like true duplicates or mergeable PR-equivalents.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45071 is the broadest central compatibility-regression report in the set; several other items are narrower v5 loading/save/import failures, so this is the best single issue to anchor the cluster if one must be chosen.", + "summary": "These issues are mostly unrelated regressions and model-specific bugs across loading, save/load, quantization, generation, and processors. None of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45092 is the clearest, most self-contained report: it describes a concrete v5 meta-initialization incompatibility with old remote-code InternVL2 checkpoints and is more specific than the other candidates.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45071 is the strongest global representative because it targets a general v5 breakage affecting config/model instantiation, which is closer to the shared pattern than the more model-specific reports.", + "best_issue_reason": "issue:45092 is the strongest representative issue because it is specific, reproducible, and points to a concrete loading-path regression rather than a broad symptom or follow-on effect.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:41628", - "right": "issue:42757", - "accept": false, - "reason": "Both are import errors, but they concern different symbols and different packages (`transformers` vs `huggingface_hub`), so they are unrelated failures." - }, - { - "left": "issue:44297", - "right": "issue:45310", - "accept": false, - "reason": "One is about `tokenizer_class` metadata saved by `save_pretrained`, the other is a `from_pretrained` loading error for Qwen3.5 MoE; same area, different bug." - }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "`PretrainedConfig` type checking in v5 is a generic regression, while this is a specific `LlamaConfig` validation rule for `head_dim`; not the same underlying issue." + "reason": "Both are config-related, but one is a PretrainedConfig type-checking regression and the other is a LlamaConfig head_dim validation error; different code paths and failure modes." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "`AutoTokenizer.from_pretrained` loading failure is unrelated to Qwen3.5 flash-attention generation behavior." + "reason": "Tokenizer loading in v5 vs Qwen3.5 flash-attention generation failure are distinct problems affecting different subsystems." }, { - "left": "issue:43504", - "right": "issue:47059", + "left": "issue:44222", + "right": "issue:44484", "accept": false, - "reason": "One is a BEiT pretrained-model legacy-field load bug; the other is a CLI breakage from `huggingface_hub` 1.22.0. Different subsystems and causes." + "reason": "One is about FP8 save_pretrained for MoE models, the other about max_shard_size default behavior; both touch saving but not the same bug." }, { - "left": "issue:44222", - "right": "issue:44484", + "left": "issue:43504", + "right": "issue:47059", "accept": false, - "reason": "FP8 `save_pretrained` behavior and the 50GB shard-size default are unrelated save-path issues." + "reason": "Legacy BEiT preset loading and a CLI incompatibility with huggingface_hub are unrelated." }, { "left": "issue:46614", "right": "issue:46932", "accept": false, - "reason": "Stale phonemizer backend state is an audio tokenizer bug, while #46932 is about Gemma4 dtype casting for quantized vision/audio embedders." + "reason": "Stale phonemizer language state in Wav2Vec2PhonemeCTCTokenizer is unrelated to Gemma4 quantized embedder dtype casting." }, { "left": "issue:43122", "right": "issue:46209", "accept": false, - "reason": "Tokenizer output differences across versions are not the same as FP8 inference failure on B200 hardware." + "reason": "Tokenization drift across versions and FP8 inference failure on B200 are different issues with no shared code-path evidence." }, { "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "A `torch.split` return-value issue in a DSA indexer is unrelated to v5 meta-initialization incompatibility for remote-code checkpoints." + "reason": "torch.split return handling in a MoE indexer is unrelated to remote-code checkpoint/meta-init incompatibility." }, { "left": "issue:44295", "right": "issue:44623", "accept": false, - "reason": "Reading `position_ids` after buffer registration is not the same as missing files from `processor.save_pretrained`." + "reason": "Reading position_ids after buffer registration and processor.save_pretrained missing files are separate serialization/attribute problems." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "An indexer split/shape issue is not a docstring problem for `position_ids`." + "reason": "A runtime torch.split behavior bug is not the same as a wrong docstring about position_ids." }, { "left": "issue:43232", "right": "issue:46424", "accept": false, - "reason": "Generation kwargs update after `sync_gpus` and cache-length handling are separate generation/cache behaviors." + "reason": "Generation kwargs syncing after sync_gpus and cache max_cache_length semantics are distinct generation/cache issues." }, { "left": "issue:43232", "right": "issue:45834", "accept": false, - "reason": "`_update_model_kwargs_for_generation` and Kosmos2.5 OCR index overflow are unrelated code paths." + "reason": "Generation-state update logic and Kosmos2.5 OCR index overflow are unrelated." }, { "left": "issue:46209", "right": "issue:46431", "accept": false, - "reason": "FP8 hardware inference failure is unrelated to processor chat-template availability." + "reason": "FP8 inference on B200 and missing chat template handling are different failure modes in different pathways." }, { "left": "issue:30990", "right": "issue:38175", "accept": false, - "reason": "Sentence Transformers loading hang and SigLIP2 zero probabilities are different model issues with different symptoms." + "reason": "Sentence Transformers loading stall and siglip2 zero-probabilities are different model/runtime bugs." }, { "left": "issue:44190", "right": "issue:46082", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to a `LlamaConfig` head-dim validation error." + "reason": "Local dataset loading in an example script is unrelated to LlamaConfig head_dim validation." }, { - "left": "issue:43475", - "right": "issue:46305", + "left": "issue:44263", + "right": "issue:44360", "accept": false, - "reason": "SAM 3 Video missing an attribute and MoonshineStreaming off-by-one masking are different model-specific bugs." + "reason": "A torch.split return-value issue and a missing ReLU in the DSA indexer are not the same bug." }, { - "left": "issue:44263", - "right": "issue:44360", + "left": "issue:43475", + "right": "issue:46305", "accept": false, - "reason": "`torch.split` return values and a missing ReLU in the DSA indexer are distinct implementation issues." + "reason": "Missing fpn_position_embeddings in SAM3 video output and a sliding-window off-by-one are different SAM/video bugs." }, { "left": "issue:46305", "right": "issue:46655", "accept": false, - "reason": "MoonshineStreaming sliding-window masking and SmolVLM video padding are separate video-processing bugs." + "reason": "Both affect video processing, but one is a sliding-window mask off-by-one and the other is a padding calculation error; not the same defect." }, { "left": "issue:46369", "right": "issue:46909", "accept": false, - "reason": "SAM3 LiteText memory overflow and a broken documentation example are unrelated." + "reason": "Memory overflow in sam3-litetext and a broken DIA documentation example are unrelated." }, { "left": "issue:43986", "right": "issue:46726", "accept": false, - "reason": "Both involve processor/vision backend setup, but one is a crash without torchvision and the other is an explicit unrecognized image processor when a backend is missing; not the same bug." + "reason": "Both mention AutoProcessor/backend availability, but one is a crash for video models without torchvision and the other is an unrecognized image processor error; not enough to be the same underlying bug." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "FineGrainedFP8 accuracy drift and Qwen3.5 MoE from_pretrained loading error are different FP8/model-loading problems." + "reason": "FineGrainedFP8 accuracy problems and Qwen3.5 from_pretrained errors are separate issues." }, { "left": "issue:22355", "right": "issue:27957", "accept": false, - "reason": "Missing `transformers.onnx` module and XLM-Roberta Flash Attention 2 are completely unrelated." + "reason": "Missing transformers.onnx module import and XLMRoberta Flash Attention 2 behavior are unrelated." }, { "left": "issue:42907", "right": "issue:46050", "accept": false, - "reason": "Saving dequantized Ministral/Devstral models and sinq-quantized models breaking after round-trip save/load are different quantization persistence bugs." + "reason": "Dequantized Ministrals/Devstrals save failures and sinq save/load breakage are both quantization-related but refer to different mechanisms and failure modes." }, { - "left": "issue:45092", - "right": "issue:45357", + "left": "issue:44279", + "right": "issue:46328", "accept": false, - "reason": "v5 meta-initialization incompatibility for remote-code checkpoints is not the same as incorrect visual encoder keys emitted by `save_pretrained`." + "reason": "A general dependency issue is not the same as SequenceFeatureExtractor.pad converting arrays inefficiently." }, { - "left": "issue:44279", - "right": "issue:46328", + "left": "issue:45092", + "right": "issue:45357", "accept": false, - "reason": "A generic dependency issue with transformers is unrelated to `SequenceFeatureExtractor.pad` inefficiency." + "reason": "Both involve Qwen3.5-family checkpoints, but one is load-time incompatibility with meta initialization while the other is save_pretrained writing incorrect visual encoder keys." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "FineGrainedFP8 accuracy loss and Qwen3-VL-Moe v5 bug are separate issues, even though both mention model internals." + "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe from_pretrained/import behavior are different bugs." }, { "left": "issue:44261", "right": "issue:45092", "accept": false, - "reason": "Missing `rms_norm_eps` in MLA q_a_layernorm causing precision differences is not the same as remote-code checkpoint incompatibility in v5." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm causing precision error is a different model/config issue than InternVL2 v5 meta-init incompatibility." }, { "left": "issue:41628", "right": "issue:43824", "accept": false, - "reason": "Importing `AutoImageProcessor` and importing `Qwen2_5_VLForConditionalGeneration` are different missing-symbol reports." + "reason": "Two import errors for different symbols/models are not evidence of the same underlying bug." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Degenerate training with flash_attention_3 and NaNs in classification heads after checkout/init are distinct training stability issues." } ] }, @@ -15580,6 +15573,7 @@ "issue:33357", "issue:39379", "issue:39401", + "issue:42371", "issue:42757", "issue:42796", "issue:42831", @@ -15598,13 +15592,13 @@ "issue:43976", "issue:43986", "issue:44188", + "issue:44263", "issue:44295", "issue:44297", "issue:44327", "issue:44373", "issue:44483", "issue:44568", - "issue:44829", "issue:44945", "issue:44964", "issue:45278", @@ -15622,238 +15616,238 @@ "issue:47217" ], "soft_pairs": [ - "issue:44829|issue:46002", - "issue:45381|issue:46305", "issue:43408|issue:46402", "issue:44964|issue:47059", + "issue:45381|issue:46305", "issue:39379|issue:46172", "issue:44483|issue:45406", "issue:42831|issue:45753", - "issue:43644|issue:46305", "issue:22355|issue:28218", + "issue:43644|issue:46305", "issue:44327|issue:47059", "issue:43986|issue:47217", "issue:44188|issue:44373", "issue:39401|issue:44568", "issue:42757|issue:45278", "issue:43408|issue:46369", - "issue:42886|issue:43066", "issue:42886|issue:42994", + "issue:42886|issue:43066", "issue:27957|issue:30990", "issue:42796|issue:43122", "issue:43232|issue:45593", "issue:45381|issue:46909", "issue:43540|issue:46909", "issue:43646|issue:46002", - "issue:42907|issue:44945", "issue:44295|issue:44297", + "issue:42907|issue:44945", "issue:30990|issue:33357", "issue:43576|issue:43784", "issue:42907|issue:44964", - "issue:43976|issue:44945" + "issue:43976|issue:44945", + "issue:42371|issue:44263" ], - "prepared_review_unit_hash": "582c190645a68b9b0c621be00ff6afb1bffc7c9e465c25b16a3bb0187ebc079c", + "prepared_review_unit_hash": "3d1b2211c8a4ab2a7ed7e9297be5c42a2fabcbeaf3a2a69501ef7e799a3b9438", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35361, - "estimated_input_tokens": 8841, - "estimated_eval_tokens": 17938 + "serialized_chars": 35803, + "estimated_input_tokens": 8951, + "estimated_eval_tokens": 18158 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are a heterogeneous mix of unrelated Transformers issues (tokenizer quirks, model-loading failures, multimodal/video bugs, quantization, CLI/import regressions). None of the soft pairs look like the same underlying bug, so there is no clear canonical duplicate in this set.", - "confidence": 0.82, - "canonical_issue_reason": null, + "summary": "This cluster is a loose mix of unrelated Transformers issues, mostly tokenizer/model regressions and version-specific breakages; there are no convincing duplicate pairs among the soft edges.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the broadest representative here: it describes a general post-upgrade regression with many import errors, which best matches the cluster\u2019s overall theme of Transformers version breakage. The rest are too specific and heterogeneous to serve as a canonical duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45278 is the best single issue representative because it is the most general and centrally framed around upgrade-related breakage, while the others are narrower and mostly unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44829", - "right": "issue:46002", - "accept": false, - "reason": "Both concern training instability, but one is about flash_attention_3 causing degenerate loss behavior and the other about NaNs after checkout/init; different failure modes and likely different fixes." - }, - { - "left": "issue:45381", - "right": "issue:46305", - "accept": false, - "reason": "Different models and code paths: qwen2.5-vl vision_position_ids vs MoonshineStreaming sliding_window_mask_function off-by-one." - }, { "left": "issue:43408", "right": "issue:46402", "accept": false, - "reason": "One is a sam3 model-type warning, the other is text encoder weights not loading for Sam3Model; not the same bug." + "reason": "Both involve SAM3, but one is a model-type warning during loading and the other is missing text-encoder weights; different concrete failures." }, { "left": "issue:44964", "right": "issue:47059", "accept": false, - "reason": "A model-loading failure for Phi-4 multimodal is unrelated to a CLI breakage caused by huggingface_hub/_add_completion." + "reason": "One is a multimodal model load failure, the other is a Hugging Face Hub CLI attribute error; unrelated code paths." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Both mention masking/position behavior, but they are different models and different bugs: Qwen2.5-VL video positions vs Moonshine streaming window off-by-one." }, { "left": "issue:39379", "right": "issue:46172", "accept": false, - "reason": "mllama masked-out mask handling and deepseek-v4 transpose logic are distinct bugs in different subsystems." + "reason": "mllama mask handling and deepseek-v4 tensor transpose are unrelated issues." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "Chat-completions request rejection and Gemma4Processor missing _tokenizer are unrelated symptoms." + "reason": "A chat/completions request validation issue is not the same as Gemma4Processor missing _tokenizer." }, { "left": "issue:42831", "right": "issue:45753", "accept": false, - "reason": "FineGrainedFP8 accuracy degradation is not the same as Qwen3_5 looping on a specific image." + "reason": "FineGrainedFP8 accuracy regression and a Qwen3_5 image infinite loop are distinct bugs." }, { - "left": "issue:43644", - "right": "issue:46305", + "left": "issue:22355", + "right": "issue:28218", "accept": false, - "reason": "Non-persistent buffer junk values vs a sliding-window attention mask off-by-one are different implementation problems." + "reason": "Module import failure for transformers.onnx and tokenizer extra-space behavior are unrelated." }, { - "left": "issue:22355", - "right": "issue:28218", + "left": "issue:43644", + "right": "issue:46305", "accept": false, - "reason": "Importing transformers.onnx and tokenizer spacing after added tokens are unrelated issues." + "reason": "Non-persistent buffer junk values and Moonshine sliding-window masking are different code paths and symptoms." }, { "left": "issue:44327", "right": "issue:47059", "accept": false, - "reason": "QA decode_spans top_k bounds error and a CLI group attribute error are unrelated." + "reason": "QA top-k bounds crash and CLI completion attribute error are unrelated." }, { "left": "issue:43986", "right": "issue:47217", "accept": false, - "reason": "AutoProcessor without torchvision crashing and VL image-placeholder handling in user text are different edge cases." + "reason": "Both are vision-language related, but one is a missing torchvision crash in AutoProcessor and the other is a placeholder-handling bug in VL models." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "torch.compile attention-kernel branching and a wrong position_ids docstring do not describe the same bug." + "reason": "A torch.compile attention-kernel divergence and a wrong docstring for position_ids are not the same bug." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer issues, but one is offset_mapping for Qwen3 and the other is add_special_tokens/BOS-EOS behavior for mdeberta-v3; too different to merge." + "reason": "Both are tokenizer bugs, but they affect different models and different behaviors: offset mapping vs BOS/EOS insertion." }, { "left": "issue:42757", "right": "issue:45278", "accept": false, - "reason": "A missing import from huggingface_hub and broad import errors after a transformers upgrade are not the same underlying defect." + "reason": "Both are import-related regressions, but one is specifically a missing huggingface_hub symbol while the other is broad upgrade breakage; not the same underlying issue." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "Model-type warning for sam3_tracker vs memory overflow in sam3-litetext are unrelated." + "reason": "Model-type mismatch warning and memory overflow in sam3-litetext are different problems." }, { "left": "issue:42886", - "right": "issue:43066", + "right": "issue:42994", "accept": false, - "reason": "Offline tokenizer cache loading and wrong tokenizer decoder type in v5 are different bugs." + "reason": "Offline tokenizer cache loading and quantized model saving failure are unrelated." }, { "left": "issue:42886", - "right": "issue:42994", + "right": "issue:43066", "accept": false, - "reason": "Offline cache loading failure is unrelated to quantized model saving failures." + "reason": "Offline cache loading and wrong tokenizer decoder type are different tokenizer/version issues, not one concrete bug." }, { "left": "issue:27957", "right": "issue:30990", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Sentence Transformers loading stuck are different problems." + "reason": "Flash Attention 2 support for XLMRoberta and Sentence-Transformers loading hang are unrelated." }, { "left": "issue:42796", "right": "issue:43122", "accept": false, - "reason": "Ministral BPE marker handling and version-to-version tokenization differences are related to tokenizers but not the same concrete bug." + "reason": "Both concern tokenization, but one is model-specific BPE marker handling and the other is version-to-version tokenization drift; not the same bug." }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Generation kwargs update after sync_gpus and missing auxiliary losses when denoising is off are unrelated." + "reason": "Generation kwargs handling after sync_gpus and D-FINE auxiliary-loss behavior are unrelated." }, { "left": "issue:45381", "right": "issue:46909", "accept": false, - "reason": "A video vision_position_ids bug and a broken DIA documentation example are unrelated." + "reason": "A Qwen2.5-VL video position bug and a broken DIA documentation example are unrelated." }, { "left": "issue:43540", "right": "issue:46909", "accept": false, - "reason": "Qwen3OmniMoe video processing ValueError is unrelated to a broken documentation example." + "reason": "Qwen3OmniMoe video processing failure and a DIA docs issue are unrelated." }, { - "left": "issue:43646", + "left": "issue:43644", "right": "issue:46002", "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are different symptoms and likely different fixes." + "reason": "Buffer initialization junk values and NaNs in classification heads are different failure modes." }, { - "left": "issue:42907", - "right": "issue:44945", + "left": "issue:44295", + "right": "issue:44297", "accept": false, - "reason": "Failing to save dequantized models and incorrect output under pipeline parallelism are unrelated." + "reason": "A position_ids buffer access error and tokenizer_config class mismatch are unrelated." }, { - "left": "issue:44295", - "right": "issue:44297", + "left": "issue:42907", + "right": "issue:44945", "accept": false, - "reason": "Reading position_ids after buffer registration and tokenizer_class mismatch in tokenizer_config.json are separate serialization/config issues." + "reason": "Quantized model saving failure and pipeline-parallelism output corruption are different code paths." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence-Transformers loading hangs and a MacOS bus error on a CLIP model are not the same bug." + "reason": "A loading hang in Sentence-Transformers and a MacOS bus error in a CLIP model are not the same issue." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "A broken transformers env command and a sentence-transformers import NameError are unrelated." + "reason": "Both are v5-related breakages, but the CLI/env command regression and the sentence-transformers import NameError are distinct symptoms and subsystems." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Dequantized model saving failure and inability to load Phi-4 multimodal are different model lifecycle problems." + "reason": "Dequantized model saving and Phi-4 multimodal loading are unrelated." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python-version compatibility failure and pipeline-parallelism output corruption are unrelated." + "reason": "Python version compatibility and pipeline-parallelism output correctness are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API guidance and a torch.split return-value bug have no shared underlying cause." } ] }, @@ -15883,7 +15877,7 @@ "issue:36296", "issue:39401", "issue:41628", - "issue:42371", + "issue:42175", "issue:42757", "issue:42907", "issue:43116", @@ -15901,7 +15895,6 @@ "issue:44060", "issue:44117", "issue:44222", - "issue:44263", "issue:44295", "issue:44297", "issue:44351", @@ -15913,6 +15906,7 @@ "issue:44704", "issue:44779", "issue:44829", + "issue:44938", "issue:44945", "issue:45200", "issue:45216", @@ -15927,13 +15921,12 @@ "issue:47059" ], "soft_pairs": [ - "issue:42371|issue:44263", "issue:41628|issue:45278", - "issue:43756|issue:44060", "issue:42907|issue:43937", + "issue:43756|issue:44060", "issue:43784|issue:45600", - "issue:44117|issue:44704", "issue:43329|issue:47059", + "issue:44117|issue:44704", "issue:39401|issue:44488", "issue:44625|issue:44829", "issue:39401|issue:44521", @@ -15944,228 +15937,229 @@ "issue:44038|issue:44297", "issue:43122|issue:46326", "issue:43122|issue:44521", - "issue:42907|issue:43452", "issue:44038|issue:44295", + "issue:42907|issue:43452", "issue:22355|issue:29159", "issue:44038|issue:44222", "issue:45216|issue:45478", "issue:43824|issue:44351", - "issue:43116|issue:45200", "issue:43122|issue:44779", + "issue:43116|issue:45200", "issue:39401|issue:46614", "issue:42757|issue:44351", "issue:36296|issue:43116", + "issue:22355|issue:29127", "issue:43278|issue:44945", - "issue:22355|issue:29127" + "issue:42175|issue:44938" ], - "prepared_review_unit_hash": "92e1e48af13e4ca35ef89b9e9825a95c5bee50dd9e7a280b20d5f9f42d50b0ad", + "prepared_review_unit_hash": "c459ad399c091aa98604b45e439674c2ee4468f2494c29cf8bc4addc046c0789", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36536, - "estimated_input_tokens": 9134, - "estimated_eval_tokens": 18524 + "serialized_chars": 36548, + "estimated_input_tokens": 9137, + "estimated_eval_tokens": 18530 }, "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 set is heterogeneous, with a few tokenizer/import-regression themes but no pair that is close enough to merge safely.", - "confidence": 0.95, - "canonical_issue_reason": "If one issue must stand in as an umbrella, issue:45278 is the broadest and most representative report because it summarizes multiple post-upgrade import failures; however, the cluster is too mixed for a true canonical duplicate.", + "summary": "No soft pair looks like a true duplicate; the set is a loose mix of unrelated Transformers regressions with some broad import/tokenizer/version-similarity overlap. I\u2019d use the broad import-regression report as the best issue representative.", + "confidence": 0.83, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: multiple import failures after upgrading Transformers, which makes it the most representative issue for the cluster\u2019s import-regression subtheme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best global representative by breadth and framing, but only as a loose umbrella for import regressions rather than a true duplicate hub.", + "best_issue_reason": "issue:45278 is the most globally useful representative because it captures a wide class of breakages after upgrade, whereas the other issues are narrower and model- or API-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42371", - "right": "issue:44263", - "accept": false, - "reason": "Unrelated: TF32 settings guidance vs a GlmMoeDsaIndexer torch.split issue." - }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "Both are import-related, but 41628 is one missing symbol while 45278 is a broader multi-import regression after upgrade." + "reason": "Both are import-related, but one is AutoImageProcessor and the other is a broad post-upgrade import regression; not the same bug." }, { - "left": "issue:43756", - "right": "issue:44060", + "left": "issue:42907", + "right": "issue:43937", "accept": false, - "reason": "Different models and different bugs: RoPE layer count vs tied-weights warning." + "reason": "Different model paths and symptoms: save/dequantization failure vs invalid GenerationConfig." }, { - "left": "issue:42907", - "right": "issue:43937", + "left": "issue:43756", + "right": "issue:44060", "accept": false, - "reason": "Save/dequantization failure and GenerationConfig validation are separate code paths." + "reason": "Unrelated model issues: RoPE layer drop vs incorrect tied-weights warning." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Import-time NameError in sentence-transformers is unrelated to stale auto_mappings config references." + "reason": "Both affect imports, but one is a missing nn symbol and the other is stale auto_mappings config references." }, { - "left": "issue:44117", - "right": "issue:44704", + "left": "issue:43329", + "right": "issue:47059", "accept": false, - "reason": "Tokenizer mapping fallback bug vs AutoProcessor kwargs forwarding; different functions and symptoms." + "reason": "Completely different code paths: multimodal token counting vs CLI completion breakage from huggingface_hub." }, { - "left": "issue:43329", - "right": "issue:47059", + "left": "issue:44117", + "right": "issue:44704", "accept": false, - "reason": "Multimodal token counting bug vs CLI completion attribute error; no shared failure mode." + "reason": "Tokenizer mapping fallback bug is unrelated to AutoProcessor not forwarding kwargs to cached_file." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Different tokenizer regressions on different models; no same underlying bug." + "reason": "Qwen3 offset_mapping bug is not the same as a model-load failure for cjvt/sleng-bert." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "num_labels propagation and flash_attention_3 training degeneration are unrelated." + "reason": "Config propagation bug for num_labels is unrelated to flash_attention_3 training degeneration." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Offset mapping bug and all-zero assistant masks are distinct tokenizer/chat-template issues." + "reason": "Different issues in tokenization/chat templating versus offset mapping." }, { "left": "issue:46172", "right": "issue:46762", "accept": false, - "reason": "Both touch numerical behavior, but on different model implementations and different code paths." + "reason": "Both mention numerical behavior, but they concern different models and different attention implementations." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "DSA indexer activation discussion vs GLM-5 RoPE implementation; not the same bug." + "reason": "DSA indexer activation omission and GLM-5 RoPE implementation are unrelated." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "AutoProcessor cache kwargs forwarding is unrelated to chat template crashes on tool-call messages." + "reason": "Different chat/template bugs: missing kwargs forwarding vs assistant message crash with tool calls." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "Same SAM3 family, but one is a model-type warning and the other is a missing attribute in the vision encoder." + "reason": "Same family, but warning about wrong model type is not the same as a missing fpn_position_embeddings attribute." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "Import/regression vs tokenizer save_pretrained mismatch; different failures." + "reason": "Qwen3-VL-Moe load issue and tokenizer save metadata mismatch are different problems." }, { "left": "issue:43122", "right": "issue:46326", "accept": false, - "reason": "Both involve tokenization behavior, but one is a broad v5 regression and the other is a mistral_common continue_final_message bug." + "reason": "Both are tokenizer-related regressions, but they affect different tokenizer implementations and failure modes." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Generic tokenizer regression vs multimodal assistant mask bug; not the same concrete issue." + "reason": "Broad tokenizer regression is not the same as multimodal assistant mask handling." }, { - "left": "issue:42907", - "right": "issue:43452", + "left": "issue:44038", + "right": "issue:44295", "accept": false, - "reason": "Saving dequantized models and gguf_file loading are separate serialization/load paths." + "reason": "MoE/version issue and position_ids buffer-read error are unrelated." }, { - "left": "issue:44038", - "right": "issue:44295", + "left": "issue:42907", + "right": "issue:43452", "accept": false, - "reason": "Unrelated: model import/regression vs reading a position_ids buffer." + "reason": "Different loading/saving code paths and different symptoms; not one underlying bug." }, { "left": "issue:22355", "right": "issue:29159", "accept": false, - "reason": "onnx import failure vs slow/fast tokenizer inconsistency are different problems." + "reason": "ONNX module import error is unrelated to slow/fast tokenizer inconsistency." }, { "left": "issue:44038", "right": "issue:44222", "accept": false, - "reason": "Qwen3-VL-Moe import/version bug vs FP8 save_pretrained issue; no shared code-path." + "reason": "Both involve MoE, but one is a model/version load issue and the other is FP8 save_pretrained behavior." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "Same model family, but save_pretrained checkpoint corruption and from_pretrained error are different directions and not clearly one fix." + "reason": "Both mention Qwen3.5 and recent versions, but one is checkpoint saving correctness and the other is from_pretrained loading error." }, { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "Different missing imports: Qwen2_5_VLForConditionalGeneration vs HybridCache." + "reason": "Distinct missing-symbol import errors for different APIs." }, { - "left": "issue:43116", - "right": "issue:45200", + "left": "issue:43122", + "right": "issue:44779", "accept": false, - "reason": "Classification example script behavior vs Gemma 4 mm_token_type_ids defaulting; unrelated." + "reason": "Two different tokenizer regressions with different models and outputs." }, { - "left": "issue:43122", - "right": "issue:44779", + "left": "issue:43116", + "right": "issue:45200", "accept": false, - "reason": "Different tokenizer regressions affecting different models and likely different root causes." + "reason": "Example-script classification output bug and Gemma 4 multimodal token-id requirement are unrelated." }, { "left": "issue:39401", "right": "issue:46614", "accept": false, - "reason": "Qwen3 offset mapping bug and stale phonemizer backend bug are unrelated tokenizer issues." + "reason": "Different tokenizer/phonemizer subsystems and different state bugs." }, { "left": "issue:42757", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but from different libraries and different missing symbols." + "reason": "Both are import errors, but they concern different missing names from different packages/modules." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training bug and multi-label classification example failure are unrelated." + "reason": "Tensor parallel training bug and empty classification outputs are not the same issue." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "No module named transformers.onnx is unrelated to LayoutLMv3 error messaging." }, { "left": "issue:43278", "right": "issue:44945", "accept": false, - "reason": "Dtype drift in evaluation is separate from incorrect outputs under pipeline parallelism." + "reason": "Embedding dtype drift and incorrect LLM output under pipeline parallelism are different failures." }, { - "left": "issue:22355", - "right": "issue:29127", + "left": "issue:42175", + "right": "issue:44938", "accept": false, - "reason": "Missing transformers.onnx import and LayoutLMv3 error-message clarity are unrelated." + "reason": "Backend dependency packaging issue is unrelated to Python 3.14 import/load failure." } ] }, @@ -16192,7 +16186,6 @@ "issue:28348", "issue:38617", "issue:39379", - "issue:42175", "issue:42371", "issue:42673", "issue:42898", @@ -16202,6 +16195,7 @@ "issue:43493", "issue:43611", "issue:43618", + "issue:43825", "issue:43874", "issue:43937", "issue:43976", @@ -16217,7 +16211,6 @@ "issue:44704", "issue:44779", "issue:44936", - "issue:44938", "issue:45072", "issue:45083", "issue:45278", @@ -16230,6 +16223,7 @@ "issue:46328", "issue:46614", "issue:46640", + "issue:46710", "issue:46726", "issue:46829", "issue:46877", @@ -16239,7 +16233,6 @@ "issue:47217" ], "soft_pairs": [ - "issue:42175|issue:44938", "issue:39379|issue:43232", "issue:43493|issue:45753", "issue:44188|issue:46328", @@ -16249,8 +16242,8 @@ "issue:38617|issue:45278", "issue:44483|issue:47217", "issue:43976|issue:45446", - "issue:43611|issue:46002", "issue:44038|issue:45478", + "issue:43611|issue:46002", "issue:43066|issue:44779", "issue:42673|issue:45072", "issue:45357|issue:45478", @@ -16265,198 +16258,206 @@ "issue:46726|issue:46981", "issue:46328|issue:47116", "issue:28348|issue:46640", - "issue:43992|issue:44704" + "issue:43992|issue:44704", + "issue:43066|issue:46710", + "issue:43116|issue:43825" ], - "prepared_review_unit_hash": "e258c089e4370dc044c457ac1f365b0912938f3c7ee325d366331ad0f30eec48", + "prepared_review_unit_hash": "03398c13ef429135d5bdac557a04b2ab6d3f13ec0d4e3cf374b7b32bc0d477f8", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35377, - "estimated_input_tokens": 8845, - "estimated_eval_tokens": 17946 + "soft_pair_count": 28, + "serialized_chars": 35754, + "estimated_input_tokens": 8939, + "estimated_eval_tokens": 18134 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose grab-bag of Transformers regressions and compatibility bugs, with soft similarities mostly coming from shared subsystems or version-v5 wording rather than the same underlying defect. I would not merge any of the proposed soft edges.", - "confidence": 0.34, - "canonical_issue_reason": "issue:45278 is the broadest, most general regression report in the set, but there is no strong evidence that this cluster is a single duplicate family; it is only the least-specific candidate.", + "summary": "No soft-edge pair looks like a true duplicate or the same fix; most are broad-theme similarities only (e.g. v5 regressions, AutoProcessor, Qwen3, flash-attention) rather than the same concrete bug. I\u2019d use the broad import-regression report as the best canonical issue in this set.", + "confidence": 0.89, + "canonical_issue_reason": "issue:45278 is the strongest umbrella-style bug report here: it describes a broad, user-visible regression after upgrading Transformers and can subsume related import-breakage reports better than the narrower model-specific tickets.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most plausible representative issue because it is broad and user-facing, but it should be treated as a loose umbrella at best, not a true duplicate canonical.", + "best_issue_reason": "Best representative issue because it is broad, clearly regression-oriented, and most suitable as a top-level reference point for related version-breakage reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42175", - "right": "issue:44938", - "accept": false, - "reason": "Both are import/runtime problems, but one is about missing TensorFlow extras in the torch install and the other is a Python 3.14 load failure; different bugs." - }, { "left": "issue:39379", "right": "issue:43232", "accept": false, - "reason": "mllama mask handling and generation kwargs after sync_gpus are unrelated code paths." + "reason": "Different subsystems and failure modes: mllama masking logic vs generation kwargs after sync_gpus." }, { "left": "issue:43493", "right": "issue:45753", "accept": false, - "reason": "SigLIP2 HF/JAX discrepancy and Qwen3_5 image-triggered looping are different model-specific bugs." + "reason": "Both involve Qwen/SigLIP-family models but describe different bugs: cross-implementation discrepancy vs an infinite loop on a specific image." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "torch.compile attention-kernel branching and SequenceFeatureExtractor.pad performance are not the same defect." + "reason": "Attention-kernel branching under torch.compile is unrelated to SequenceFeatureExtractor.pad conversion overhead." }, { "left": "issue:46231", "right": "issue:46877", "accept": false, - "reason": "A docs parameter-name mismatch is not the same as auto_docstring emitting lint output at import time." + "reason": "Documentation parameter-name mismatch is not the same as auto_docstring printing diagnostics at import time." }, { "left": "issue:42898", "right": "issue:46614", "accept": false, - "reason": "Tokenizer cleanup behavior in v5 and stale espeak backend selection are unrelated tokenization/speech issues." + "reason": "Tokenizer cleanup behavior in v5 and a stale phonemizer backend are separate tokenization/phonemization issues." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "GLM-5 GenerationConfig validation and GLM-5 RoPE implementation are different bugs in different parts of the model." + "reason": "GenerationConfig validation and GLM-5 RoPE implementation are different concrete problems." }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "One is a missing symbol import, the other is a broad upgrade-time import-breakage report; same theme, not the same concrete bug." + "reason": "Related release-breakage theme, but one is a specific missing symbol import and the other is a broad multi-import regression; not enough evidence they are the same underlying bug." }, { "left": "issue:44483", "right": "issue:47217", "accept": false, - "reason": "Chat/completions API regression and vision-language placeholder handling are different request-path problems." + "reason": "Chat/completions request handling and VL image-placeholder handling are not the same bug." }, { "left": "issue:43976", "right": "issue:45446", "accept": false, - "reason": "Python version gating and a PyTorch version check for flex_attention are separate compatibility checks." + "reason": "Python-version compatibility issues do not match a flex_attention PyTorch version guard bug." }, { - "left": "issue:43611", - "right": "issue:46002", + "left": "issue:44038", + "right": "issue:45478", "accept": false, - "reason": "base_model_prefix loading regression and NaNs in classification heads are unrelated failures." + "reason": "Both mention Qwen3 MoE, but one is a Qwen3-VL-Moe issue and the other is a Qwen3.5 MoE from_pretrained error; different code paths and model variants." }, { - "left": "issue:44038", - "right": "issue:45478", + "left": "issue:43611", + "right": "issue:46002", "accept": false, - "reason": "Qwen3-VL-Moe from a bug report and Qwen3.5 Moe from_pretrained error are not the same code-path failure." + "reason": "Model-loading base_model_prefix regression and NaNs in classification heads are unrelated." }, { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Both mention tokenizer regressions in v5, but they concern different models and different tokenizer behavior." + "reason": "The tokenizer-decoder-type report may be a possible root cause, but the DeepSeek issue is not clearly the same concrete bug without more evidence." }, { "left": "issue:42673", "right": "issue:45072", "accept": false, - "reason": "VRAM leakage in multithreaded dataloading and bf16 dtype mismatches are not the same bug." + "reason": "VRAM leakage in Qwen3ForCausalLM under dataloader threads is unrelated to dtype mismatches in inference tests." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "save_pretrained writing wrong visual encoder keys is distinct from from_pretrained failing on Qwen3.5 Moe." + "reason": "Both concern Qwen3.5 MoE, but save_pretrained key serialization and from_pretrained loading errors are different failures." }, { "left": "issue:44355", "right": "issue:46231", "accept": false, - "reason": "Compiled Python file errors and a documentation parameter typo are unrelated." + "reason": "Compiled Python file execution errors are unrelated to a training API doc typo." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "Qwen3.5 num_labels propagation and NaNs in classification heads are different issues." + "reason": "num_labels propagation in Qwen3.5 config and NaNs in classification heads are different bugs." }, { "left": "issue:43116", "right": "issue:43493", "accept": false, - "reason": "The example script returning empty multi-label results is not the same as a SigLIP2 implementation discrepancy." + "reason": "Example-script multi-label empty output and SigLIP2 implementation discrepancy are unrelated." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "A feature-extractor length helper bug and a flex_attention version-check bug are unrelated." + "reason": "Feature-length helper behavior in qwen3_omni_moe is unrelated to a flex_attention import/version check." }, { "left": "issue:44610", "right": "issue:46909", "accept": false, - "reason": "Processor/model input-size mismatch and a broken documentation example are not the same underlying problem." + "reason": "Processor output size mismatch and a broken documentation example are not the same issue." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 API guidance and a missing multimodal patch-count method are unrelated." + "reason": "TF32 API guidance and a missing image-patch-count method are unrelated." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "Optimizer argument handling and trainer.evaluate failing after train are different training bugs." + "reason": "Optimizer argument handling is unrelated to trainer.evaluate() failing after train()." }, { "left": "issue:43618", "right": "issue:46829", "accept": false, - "reason": "Both are output-field mixups, but they affect different model outputs and different tensors; not enough to treat as one bug." + "reason": "Both are output-field assignment issues, but they affect different model classes and different output members; not clearly the same fix." }, { "left": "issue:46726", "right": "issue:46981", "accept": false, - "reason": "Missing vision backend handling in AutoProcessor and repeated HTTP 429 retries are completely different AutoProcessor behaviors." + "reason": "AutoProcessor backend-missing messaging and HTTP-429 retry behavior are distinct from each other." }, { "left": "issue:46328", "right": "issue:47116", "accept": false, - "reason": "Padding performance and streaming pipeline regression are different issues with no shared concrete failure mode." + "reason": "SequenceFeatureExtractor.pad performance overhead is unrelated to streaming pipeline regression." }, { "left": "issue:28348", "right": "issue:46640", "accept": false, - "reason": "A specific GPT2LMHeadModel flash-attention feature request is not the same as a generic Flash Attention 2 issue." + "reason": "Flash-attention support request for GPT2LMHeadModel is not the same as an open Flash Attention 2 bug report." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "Missing embed_tokens.weight in UMT5Encoder and AutoProcessor kwargs forwarding are unrelated." + "reason": "UMT5Encoder weight loading and AutoProcessor cached_file kwargs handling are different code paths." + }, + { + "left": "issue:43066", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer decoder-type regression and DeepSeek-R1 accuracy regression are too far apart to treat as the same concrete bug." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example classification script output and pipeline translation-task messaging are unrelated." } ] }, @@ -16487,14 +16488,11 @@ "issue:42673", "issue:42796", "issue:43064", - "issue:43066", - "issue:43116", "issue:43441", "issue:43450", "issue:43673", "issue:43688", "issue:43824", - "issue:43825", "issue:43827", "issue:43937", "issue:43986", @@ -16504,6 +16502,7 @@ "issue:44380", "issue:44410", "issue:44466", + "issue:44704", "issue:44779", "issue:44829", "issue:44977", @@ -16515,253 +16514,262 @@ "issue:45341", "issue:45405", "issue:45446", + "issue:45561", "issue:45563", "issue:45588", "issue:46002", "issue:46077", "issue:46328", + "issue:46512", "issue:46697", - "issue:46710", "issue:46752", "issue:46762", "issue:46897", "issue:46981", "issue:47000", - "issue:47059" + "issue:47059", + "issue:47110" ], "soft_pairs": [ - "issue:43066|issue:46710", - "issue:43116|issue:43825", - "issue:42673|issue:42796", "issue:43827|issue:45446", + "issue:42673|issue:42796", "issue:44995|issue:46697", "issue:45092|issue:46762", "issue:33453|issue:39401", "issue:44263|issue:46762", - "issue:43064|issue:43673", - "issue:43824|issue:46981", "issue:43673|issue:46328", + "issue:43824|issue:46981", + "issue:43064|issue:43673", "issue:39379|issue:43688", "issue:42222|issue:43441", "issue:42222|issue:47000", - "issue:44380|issue:46077", "issue:45405|issue:45446", "issue:43937|issue:45563", - "issue:42222|issue:46752", + "issue:44380|issue:46077", "issue:44117|issue:45588", + "issue:42222|issue:46752", "issue:44291|issue:45071", - "issue:39379|issue:46897", "issue:44466|issue:46002", + "issue:39379|issue:46897", "issue:44410|issue:45310", "issue:39401|issue:44779", "issue:43937|issue:44977", "issue:45071|issue:47059", - "issue:45341|issue:46328", "issue:43450|issue:43986", - "issue:44829|issue:45216" + "issue:45341|issue:46328", + "issue:44829|issue:45216", + "issue:44704|issue:47110", + "issue:45071|issue:46512", + "issue:45341|issue:45561" ], - "prepared_review_unit_hash": "9915cde79a73ac5207cf6e19b9a2b7e80469e7eebf2ce57c1d35c699d5c0587b", + "prepared_review_unit_hash": "41b9cb991216a40709c102d1694a21e27ca3226c439901cc38cf72a9a76c7564", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36393, - "estimated_input_tokens": 9099, - "estimated_eval_tokens": 18454 + "soft_pair_count": 30, + "serialized_chars": 36659, + "estimated_input_tokens": 9165, + "estimated_eval_tokens": 18586 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No duplicate clusters are supported by the evidence here; every soft pair appears to be a distinct bug or change, even when the symptoms are superficially similar (tokenizers, v5 regressions, attention backends, video models, etc.).", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the batch spans unrelated defects across different models, APIs, and subsystems, so there is no single issue that clearly represents the same underlying problem.", + "summary": "These items are a heterogeneous set of unrelated Transformers bugs, so none of the soft-edge pairs look like true duplicates or mergeable PR-equivalents. Most pairs share only a broad subsystem (tokenizers, generation, processors, attention, training) but differ in the concrete code path and failure mode.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43824 is a high-signal, user-facing bug with multiple references and a clear concrete failure mode (importing a missing Qwen2_5_VL symbol), making it a good representative issue for this cluster.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong global canonical candidate because none of the soft pairs looks like a true duplicate or the same concrete code-path failure.", + "best_issue_reason": "issue:43824 is the strongest overall representative: specific, reproducible, externally visible, and broadly impactful compared with the many more niche or lower-signal items.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43066", - "right": "issue:46710", - "accept": false, - "reason": "Both involve v5/model behavior regressions, but one is a tokenizer decoder-type problem and the other is an accuracy regression on a specific DeepSeek model; they are not the same defect." - }, - { - "left": "issue:43116", - "right": "issue:43825", + "left": "issue:43827", + "right": "issue:45446", "accept": false, - "reason": "One is a broken example script for multi-label classification, the other is a misleading pipeline error message about translation support; different code paths and symptoms." + "reason": "Docs still referencing pipeline() vs an incorrect PyTorch version gate for AuxRequest; different layers and unrelated fixes." }, { "left": "issue:42673", "right": "issue:42796", "accept": false, - "reason": "VRAM leakage in Qwen3 causal LM threads and tokenizer BPE-marker handling are unrelated bugs." - }, - { - "left": "issue:43827", - "right": "issue:45446", - "accept": false, - "reason": "Docs referencing removed pipeline APIs and a PyTorch version check in flex_attention are unrelated changes." + "reason": "Both are tokenizer/model bugs, but one is VRAM leakage in multithreaded loading and the other is BPE marker handling; different symptoms and code paths." }, { "left": "issue:44995", "right": "issue:46697", "accept": false, - "reason": "A stale cache crash in GlmMoeDsa and over-padding in Swin2SR image processing are different subsystems and failures." + "reason": "GLM-Moe stale indexer cache crash vs Swin2SR over-padding; unrelated components and failures." }, { "left": "issue:45092", "right": "issue:46762", "accept": false, - "reason": "Remote-code checkpoint/meta-init incompatibility is unrelated to numerical differences in Minimax MSA attention." + "reason": "Meta-initialization incompatibility for old remote-code checkpoints vs numerical mismatch in Minimax attention; not the same bug." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Both are tokenizer issues, but one is generic tokenizer loading regression and the other is a specific offset_mapping bug for Qwen3; not the same underlying problem." + "reason": "Both mention tokenizers, but loading regression and wrong offset mapping are different failures in different model families." }, { "left": "issue:44263", "right": "issue:46762", "accept": false, - "reason": "A torch.split return-value issue in GlmMoeDsaIndexer is unrelated to an attention implementation mismatch." + "reason": "torch.split return-value issue in GlmMoeDsaIndexer vs attention numerical differences; unrelated." }, { - "left": "issue:43064", - "right": "issue:43673", + "left": "issue:43673", + "right": "issue:46328", "accept": false, - "reason": "Optimizer-state corruption under FSDP2/PEFT and missing GenerationMixin cache in chunked prefill are separate training vs generation bugs." + "reason": "Generation cache missing during chunked prefill vs a padding performance inefficiency; not the same underlying bug." }, { "left": "issue:43824", "right": "issue:46981", "accept": false, - "reason": "ImportError for a Qwen2.5 VL class and repeated 429 retries in AutoProcessor are not the same change." + "reason": "Missing Qwen2_5_VL import symbol vs repeated HTTP 429 retries in AutoProcessor; different failure mode and fix." }, { - "left": "issue:43673", - "right": "issue:46328", + "left": "issue:43064", + "right": "issue:43673", "accept": false, - "reason": "A generation cache bug and a padding performance issue in feature extraction are unrelated." + "reason": "Optimizer-state corruption under FSDP2/PEFT loading vs GenerationMixin cache issue; different subsystems and root causes." }, { "left": "issue:39379", "right": "issue:43688", "accept": false, - "reason": "Incorrect mllama mask handling and auxiliary-loss normalization bugs affect different model components." + "reason": "mllama mask handling bug vs aux-loss normalization bug; both training-related but distinct logic." }, { "left": "issue:42222", "right": "issue:43441", "accept": false, - "reason": "A broken vitpose model import/setup and a Ministral FlashAttention failure are unrelated model-specific bugs." + "reason": "Broken vitpose models vs Ministral FlashAttention failure; unrelated model families and code paths." }, { "left": "issue:42222", "right": "issue:47000", "accept": false, - "reason": "vitpose breakage and RTDetrLoss infeasible-cost-matrix crashes do not share the same code path." - }, - { - "left": "issue:44380", - "right": "issue:46077", - "accept": false, - "reason": "GPT2 attention scaling with SDPA/FlashAttention and Gemma4 bidirectional-mask handling are both attention-related, but they are different model-specific bugs." + "reason": "vitpose breakage vs RTDetr loss infeasibility; no concrete overlap." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "A dependency-version bump issue and a PyTorch import/version-check bug are unrelated." + "reason": "PEFT minimum-version mismatch vs incorrect PyTorch version check for flex_attention; unrelated compatibility issues." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "An invalid GenerationConfig error and a stale warning about num_return_sequences are different generation issues." + "reason": "Invalid GenerationConfig vs stale warning for num_return_sequences; different generation-validation bugs." }, { - "left": "issue:42222", - "right": "issue:46752", + "left": "issue:44380", + "right": "issue:46077", "accept": false, - "reason": "vitpose breakage and empty-list apply_chat_template crashes are unrelated." + "reason": "GPT2 attention scaling ignored under SDPA/FlashAttention vs Gemma4 bidirectional mask construction; different model logic." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and a flash_attention AttributeError on s_aux=None are unrelated." + "reason": "TOKENIZER_MAPPING_NAMES returning None vs flash_attention crash on s_aux=None; unrelated lookup vs runtime-attribute failures." }, { - "left": "issue:44291", - "right": "issue:45071", + "left": "issue:42222", + "right": "issue:46752", "accept": false, - "reason": "Unexpected _is_hf_initialized init args and PretrainedConfig type-checking regressions are both v5-related, but they are distinct failure points." + "reason": "vitpose model breakage vs apply_chat_template empty-list IndexError; no shared code path." }, { - "left": "issue:39379", - "right": "issue:46897", + "left": "issue:44291", + "right": "issue:45071", "accept": false, - "reason": "These are both training/math bugs, but the mllama mask issue is not the same as Florence2's double-shift loss bug." + "reason": "Unexpected _is_hf_initialized arg in init_empty_weights vs PretrainedConfig type-checking breakage; different API regressions." }, { "left": "issue:44466", "right": "issue:46002", "accept": false, - "reason": "Inconsistent lm_head.weight serialization and NaNs in classification heads are separate save/load vs initialization issues." + "reason": "lm_head.weight serialization inconsistency vs NaNs in classification heads; distinct training/save behavior bugs." + }, + { + "left": "issue:39379", + "right": "issue:46897", + "accept": false, + "reason": "mllama masked-out mask handling vs Florence2 double-shift training-loss bug; different models and failure mechanisms." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Missing qwen3next projections and Qwen3.5 from_pretrained errors are different model-loading bugs." + "reason": "Missing qwen3next projections in layer 0 vs Qwen3.5 MoE from_pretrained error; related family, but not the same concrete bug." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Two tokenizer regressions, but for different models and likely different tokenization logic; not enough to treat as duplicates." + "reason": "Qwen3 offset_mapping regression vs Deepseek tokenizer producing incorrect results; both tokenizer bugs but not the same issue." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "GenerationConfig validation and Qwen3.5 FlashAttention generation failures are separate issues." + "reason": "Invalid GenerationConfig vs Qwen3.5 flash-attention generation failure; different bug classes." }, { "left": "issue:45071", "right": "issue:47059", "accept": false, - "reason": "PretrainedConfig type checking and CLI completion breakage in huggingface_hub are unrelated." + "reason": "PretrainedConfig type checking regression vs cli failure with huggingface_hub 1.22.0; unrelated surfaces." }, { - "left": "issue:45341", - "right": "issue:46328", + "left": "issue:43450", + "right": "issue:43986", "accept": false, - "reason": "A small testing_utils bug and a feature-extractor padding performance issue are unrelated." + "reason": "Batched video-processor output shape bug vs AutoProcessor crash without torchvision; both video-related, but different conditions and fixes." }, { - "left": "issue:43450", - "right": "issue:43986", + "left": "issue:45341", + "right": "issue:46328", "accept": false, - "reason": "Batched video processor shape errors and missing-torchvision AutoProcessor crashes are different video pipeline failures." + "reason": "testing_utils bug vs SequenceFeatureExtractor.pad inefficiency; one is correctness/race-like, the other is performance." }, { "left": "issue:44829", "right": "issue:45216", "accept": false, - "reason": "Degenerate training with flash_attention_3 and incorrect checkpoint saving in Qwen3.5 are unrelated bugs." + "reason": "FlashAttention-3 causing degenerate training vs incorrect Qwen3.5 checkpoint saving; different code paths and symptoms." + }, + { + "left": "issue:44704", + "right": "issue:47110", + "accept": false, + "reason": "AutoProcessor not forwarding kwargs to cached_file vs special_tokens list lacking keys; unrelated input-shape/API bugs." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "Type-checking regression for PretrainedConfig vs stale config.dtype after from_config(dtype=...); separate config-handling issues." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "A bug in testing_utils vs xdist workers racing on captured_info.txt; both in testing utilities, but different failures and likely different fixes." } ] }, @@ -16796,7 +16804,10 @@ "issue:43122", "issue:43208", "issue:43408", + "issue:43673", "issue:43742", + "issue:43874", + "issue:43931", "issue:43937", "issue:44075", "issue:44077", @@ -16809,7 +16820,6 @@ "issue:44485", "issue:44493", "issue:44496", - "issue:44704", "issue:44877", "issue:44933", "issue:45071", @@ -16818,29 +16828,23 @@ "issue:45310", "issue:45341", "issue:45478", - "issue:45561", "issue:45632", "issue:45656", "issue:46178", "issue:46209", "issue:46231", "issue:46298", - "issue:46512", "issue:46617", "issue:46752", "issue:46762", "issue:46894", "issue:46981", - "issue:47110", "issue:47116" ], "soft_pairs": [ - "issue:44704|issue:47110", - "issue:45071|issue:46512", - "issue:45341|issue:45561", + "issue:44117|issue:45081", "issue:43408|issue:43742", "issue:42617|issue:45341", - "issue:44117|issue:45081", "issue:45632|issue:46981", "issue:38617|issue:41628", "issue:43122|issue:46298", @@ -16861,198 +16865,215 @@ "issue:43208|issue:46894", "issue:43097|issue:44493", "issue:44077|issue:46981", - "issue:46752|issue:47116" + "issue:46752|issue:47116", + "issue:43673|issue:45341", + "issue:43931|issue:44410", + "issue:41628|issue:45071", + "issue:43874|issue:44261", + "issue:44484|issue:45478" ], - "prepared_review_unit_hash": "4059c219bbef97ab1b1280f0fd0b5d56bfa85c3f0a9925f248ab39f7ca2805ee", + "prepared_review_unit_hash": "c0d15bd4c505a13e6e78f648ecae8556863c7f50a63d464e8dec43e2d56c865b", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35281, - "estimated_input_tokens": 8821, - "estimated_eval_tokens": 17898 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35345, + "estimated_input_tokens": 8837, + "estimated_eval_tokens": 17930 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous: import/export regressions, tokenizer behavior changes, model-specific loading bugs, optimizer issues, and unrelated test/runtime problems. There is no strong duplicate set, so the safest conclusion is that these are separate issues rather than one unified bug.", - "confidence": 0.41, - "canonical_issue_reason": null, + "summary": "This is a loose cluster of Transformers bug reports, mostly v5 regression and model/tokenizer-loading issues, but the soft pairs generally point to different code paths, symbols, or models rather than true duplicates.", + "confidence": 0.77, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report ('Many import errors after update from 4.57.0 to 5.5.0') and best matches the cluster's general upgrade-regression theme, though the overall set is quite heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the broadest representative because it covers the recurring v5 import/API-breakage theme seen in a few other reports, but it is still not a true umbrella for the whole cluster.", + "best_issue_reason": "issue:45278 is the most representative anchor because it is broad and generic enough to cover the shared v5 breakage theme; it is still only an imperfect canonical since most other items are narrower and unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44704", - "right": "issue:47110", - "accept": false, - "reason": "Both are import/argument-shape problems, but one is about AutoProcessor kwargs to cached_file and the other is about malformed special_tokens input; different code paths and failures." - }, - { - "left": "issue:45071", - "right": "issue:46512", - "accept": false, - "reason": "A type-checking regression in PreTrainedConfig is unrelated to from_config(dtype=...) leaving config.dtype stale." - }, - { - "left": "issue:45341", - "right": "issue:45561", + "left": "issue:44117", + "right": "issue:45081", "accept": false, - "reason": "Both mention testing_utils, but one is a minor bug report while the other is a pytest-xdist race on captured_info.txt; not the same defect." + "reason": "Both are tokenizer-related, but one is a None mapping in from_pretrained and the other is a Mistral regex/backend_tokenizer crash; different code paths." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "A sam3 model-type warning is not the same underlying bug as a MobileLLM loading KeyError." + "reason": "Different models and failures: sam3 type warning vs MobileLLM key error." }, { "left": "issue:42617", "right": "issue:45341", "accept": false, - "reason": "3d_parallel.py runtime failure and a testing_utils bug are unrelated." - }, - { - "left": "issue:44117", - "right": "issue:45081", - "accept": false, - "reason": "Both are tokenizer-related, but one is a None mapping assumption and the other is a Mistral regex patch crash; different root causes." + "reason": "Unrelated: a runtime failure in 3d_parallel.py vs a small testing_utils bug." }, { "left": "issue:45632", "right": "issue:46981", "accept": false, - "reason": "Cache-path collisions for trust_remote_code are unrelated to repeated HTTP 429 retries in AutoProcessor.from_pretrained." + "reason": "Cache-path collision for trust_remote_code vs repeated 429 retries per config file; unrelated cache vs HTTP logic." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols and likely different API regressions." + "reason": "Both are import errors, but for different symbols/modules and likely different missing exports." }, { "left": "issue:43122", "right": "issue:46298", "accept": false, - "reason": "Tokenizer behavior changes and a noisy logging message are not the same bug." + "reason": "Tokenizer output change vs a noisy logging message; not the same bug." }, { "left": "issue:44877", "right": "issue:46231", "accept": false, - "reason": "A strict config-loading failure is unrelated to a documentation/parameter-name mismatch." + "reason": "Config-loading failure vs documentation parameter-name mismatch." }, { "left": "issue:42222", "right": "issue:46178", "accept": false, - "reason": "Broken vitpose model loading and an unsupported tensorflow_text backend error are different problems." + "reason": "Vitpose breakage vs tensorflow_text backend mapping error; different subsystems." }, { "left": "issue:39401", "right": "issue:46209", "accept": false, - "reason": "Qwen3 tokenizer offset mapping and FP8 inference on B200 are completely different issues." + "reason": "Tokenizer offset_mapping bug vs FP8 inference failure on B200." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API guidance and an MLA rms_norm_eps precision bug do not share the same code-path problem." + "reason": "TF32 API/settings discussion vs MLA layernorm epsilon precision issue." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "A GenerationConfig validation error and an unrecognized model/config error are distinct loading/validation failures." + "reason": "Invalid GenerationConfig vs unrecognized model/config type; unrelated loading errors." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "A shard-size question and a Qwen3.5 MoE from_pretrained error are unrelated." + "reason": "Save-pretrained shard sizing question vs Qwen3.5 MoE from_pretrained error." }, { "left": "issue:44261", "right": "issue:46762", "accept": false, - "reason": "Both are numerical-accuracy complaints, but they target different models and implementations; not the same underlying bug." + "reason": "Both involve numerical behavior, but one is a missing rms_norm_eps/config precision issue and the other is an attention reference mismatch; different code paths." }, { "left": "issue:44933", "right": "issue:45278", "accept": false, - "reason": "The former is a specific missing import from image_utils, while the latter is a broad post-upgrade import-regression report; related theme, not the same bug." + "reason": "Specific missing image_utils import vs broad post-upgrade import errors; too different to merge." }, { "left": "issue:42371", "right": "issue:46617", "accept": false, - "reason": "TF32 API settings and a trust_remote_code cache FileNotFoundError are unrelated." + "reason": "TF32 control settings vs FileNotFoundError in trust_remote_code cache loading." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Missing projections in qwen3next and a Qwen3.5 MoE from_pretrained error are different model-specific failures." + "reason": "Different Qwen issues: missing projections in qwen3next vs Qwen3.5 MoE load failure." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "An SGD-args-not-used bug is not the same as optimizer.step being called twice under deepspeed." + "reason": "Both optimizer-related, but one ignores SGD args while the other double-calls optimizer.step with deepspeed." }, { "left": "issue:38617", "right": "issue:44220", "accept": false, - "reason": "An import error for layer_type_validation and an audio feature extraction issue are unrelated." + "reason": "ImportError for a missing configuration utility symbol vs an audio feature extraction bug." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "A torch.split return-value bug and a RoPE implementation discussion are not the same code-path failure." + "reason": "torch.split return handling vs GLM-5 RoPE implementation; unrelated failures." }, { "left": "issue:43066", "right": "issue:43122", "accept": false, - "reason": "Both involve tokenizer changes in v5, but one is about decoder type and the other about differing tokenization outputs; too broad to merge." + "reason": "Both are tokenizer regressions in v5, but one is decoder type selection and the other is changed tokenization output; not the same fix." }, { "left": "issue:43208", "right": "issue:46894", "accept": false, - "reason": "xLSTM training blockers and a CohereASR double-shift loss bug are unrelated model/training issues." + "reason": "Different model families and symptoms: xLSTM training blockers vs CohereASR loss double-shift." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removing tie_embeddings_and_encoder_decoder and unexpected position-id keys are not the same regression." + "reason": "tie_embeddings removal vs unexpected position-id key; unrelated API/config regressions." }, { "left": "issue:44077", "right": "issue:46981", "accept": false, - "reason": "A patchtsmixer post_init policy change is unrelated to HTTP 429 retry behavior." + "reason": "PatchTSMixer post_init policy vs repeated HTTP 429 retries; different concerns." }, { "left": "issue:46752", "right": "issue:47116", "accept": false, - "reason": "An empty-list chat-template crash and a streaming pipeline regression are separate failures." + "reason": "Empty-list chat template crash vs streaming pipeline regression; different code paths." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache issue during chunked prefill vs a testing_utils bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch vs qwen3next missing projections; different model-loading problems." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor import error vs PreTrainedConfig type-checking issue; not the same regression." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image-processor method vs MLA layernorm epsilon precision mismatch." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Shard-size question vs the Qwen3.5 MoE from_pretrained error." } ] }, @@ -17078,9 +17099,9 @@ "nodes": [ "issue:27957", "issue:38617", - "issue:41628", "issue:42757", "issue:42886", + "issue:42890", "issue:43012", "issue:43064", "issue:43097", @@ -17088,6 +17109,7 @@ "issue:43352", "issue:43441", "issue:43452", + "issue:43531", "issue:43577", "issue:43673", "issue:43874", @@ -17098,52 +17120,46 @@ "issue:44261", "issue:44263", "issue:44315", + "issue:44351", "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:45588", "issue:45753", "issue:46050", "issue:46123", "issue:46172", + "issue:46298", "issue:46369", "issue:46655", "issue:46664", + "issue:46752", "issue:46897", - "issue:46981", - "issue:47000" + "issue:46981" ], "soft_pairs": [ - "issue:43673|issue:45341", - "issue:43931|issue:44410", - "issue:41628|issue:45071", - "issue:43874|issue:44261", - "issue:44484|issue:45478", "issue:27957|issue:43352", "issue:44483|issue:45464", "issue:44263|issue:46172", "issue:44360|issue:46172", - "issue:43012|issue:43577", "issue:43441|issue:44829", "issue:43931|issue:44038", + "issue:43012|issue:43577", "issue:43097|issue:43452", "issue:44261|issue:44485", "issue:43986|issue:45341", "issue:45003|issue:45341", "issue:42886|issue:46981", - "issue:45341|issue:47000", + "issue:43531|issue:45341", "issue:45753|issue:46369", "issue:44117|issue:44987", "issue:43874|issue:44263", @@ -17157,228 +17173,219 @@ "issue:43064|issue:44315", "issue:43986|issue:45372", "issue:43208|issue:46897", - "issue:46655|issue:46664" + "issue:46655|issue:46664", + "issue:38617|issue:44351", + "issue:42890|issue:46298", + "issue:46298|issue:46752" ], - "prepared_review_unit_hash": "a0e105ca16751f75c1aaee9ccc117fb54629d6587bd963a334d2117f44e098bc", + "prepared_review_unit_hash": "522bc950efced0864059a2d784ddf60d28c91ebc698017181650b8ab5141a8ee", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37134, - "estimated_input_tokens": 9284, - "estimated_eval_tokens": 18824 + "soft_pair_count": 30, + "serialized_chars": 36561, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous transformer bug reports with a few superficial similarities; I do not see any true duplicate pair to merge. The only near-related items are the two SmolVLM video-processor reports, but they still look like distinct bugs.", - "confidence": 0.61, - "canonical_issue_reason": "issue:46655 is the closest anchor because it matches the only near-duplicate-looking pair (with issue:46664) in the otherwise heterogeneous set.", + "summary": "This cluster is heterogeneous: most items are unrelated Transformers regressions spanning model-specific loading, flash-attention, generation, video processors, and utility/import issues. None of the soft pairs look like true duplicates of the same underlying bug, so I would not merge any of them as duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "If a single anchor is needed, issue 43097 is the broadest core API regression with downstream impact and inbound references, but the cluster is not a true duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46655 is the most representative bug report here, but the cluster is too mixed to be a strong universal canonical.", + "best_issue_reason": "Issue 43097 is the strongest representative issue because it describes a central user-facing regression in core Transformers behavior, not just a model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43673", - "right": "issue:45341", - "accept": false, - "reason": "Different areas: generation cache/chunked prefill vs a testing_utils bug." - }, - { - "left": "issue:43931", - "right": "issue:44410", - "accept": false, - "reason": "Same broad Qwen/VL ecosystem, but one is a load-time weight-shape mismatch and the other is a missing projection definition." - }, - { - "left": "issue:41628", - "right": "issue:45071", - "accept": false, - "reason": "Import error for AutoImageProcessor is unrelated to PreTrainedConfig type-checking behavior." - }, - { - "left": "issue:43874", - "right": "issue:44261", - "accept": false, - "reason": "One is a missing image-processor method; the other is an MLA RMS eps precision issue." - }, - { - "left": "issue:44484", - "right": "issue:45478", - "accept": false, - "reason": "A save_pretrained shard-size question is not the same bug as a Qwen3.5 MoE from_pretrained error." - }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "Both mention Flash Attention 2, but they are different models and different failure modes." + "reason": "Both mention Flash Attention 2, but they are different model-specific support failures with different code paths and fixes." }, { "left": "issue:44483", "right": "issue:45464", "accept": false, - "reason": "Both hit chat/completions, but one is a generic API acceptance failure and the other is a streaming inference issue on a specific model." + "reason": "Both concern chat/completions, but one is a general request-acceptance regression while the other is a streaming inference failure on Qwen3.5." }, { "left": "issue:44263", "right": "issue:46172", "accept": false, - "reason": "Unrelated tensor/math bugs in different model families." + "reason": "Completely different bugs: a torch.split return-value issue versus a DeepSeek v4 dtype/transpose problem." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "DSA indexer ReLU concern does not match the DeepSeek dtype/transpose bug." - }, - { - "left": "issue:43012", - "right": "issue:43577", - "accept": false, - "reason": "Both are dtype-related, but one is a PyTorch warning during compile and the other is a loading-time float32 retention bug." + "reason": "Unrelated defects in different model internals; shared DeepSeek-style ML context is not enough for duplication." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "FlashAttention-related, but one is a model support/compatibility error and the other is a training degeneration regression." + "reason": "Both involve flash attention, but one is a model load/runtime failure and the other is degenerate training with flash_attention_3." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Related Qwen3-VL naming, but the reported failures and root causes differ." + "reason": "Different Qwen3-VL-family failures with distinct symptoms: weight-shape mismatch versus a Transformers 5.0 compatibility issue." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Both are dtype-related, but one is a compile-time warning and the other is a model loading dtype propagation bug." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "Removal of a config helper is unrelated to gguf_file breaking from_pretrained." + "reason": "A removed config/API issue is unrelated to gguf_file loading breakage in AutoTokenizer/AutoModel." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "Different model internals: MLA q_a_layernorm precision vs GLM-5 RoPE implementation." + "reason": "Different model implementation bugs: missing rms_norm_eps in MLA versus a GLM-5 RoPE implementation issue." }, { "left": "issue:43986", "right": "issue:45341", "accept": false, - "reason": "AutoProcessor crash without torchvision is unrelated to the testing_utils bug." + "reason": "No overlap beyond being bugs in the repo; one is an AutoProcessor/torchvision dependency crash, the other is a testing_utils bug." }, { "left": "issue:45003", "right": "issue:45341", "accept": false, - "reason": "Unsafe sys.modules access is a different code path from the testing_utils issue." + "reason": "Unrelated utility-layer issues; unsafe sys.modules access is not the same problem as the testing_utils bug." }, { "left": "issue:42886", "right": "issue:46981", "accept": false, - "reason": "Both are loading/retry issues, but offline cache fallback and repeated 429 retries are distinct problems." + "reason": "Both involve hub/network behavior, but offline cache loading and repeated 429 retries are different failures with different fixes." }, { - "left": "issue:45341", - "right": "issue:47000", + "left": "issue:43531", + "right": "issue:45341", "accept": false, - "reason": "Testing utility bug vs RTDetrLoss infeasible cost-matrix crash are unrelated." + "reason": "A Qwen3-MoE sliding_window bug is unrelated to a testing_utils issue." }, { "left": "issue:45753", "right": "issue:46369", "accept": false, - "reason": "An infinite-loop image-specific bug is not the same as a memory overflow in sam3-litetext." + "reason": "Different model families and symptoms: an infinite-loop image case versus a SAM3-LiteText memory overflow." }, { "left": "issue:44117", "right": "issue:44987", "accept": false, - "reason": "Both affect from_pretrained, but tokenizer mapping None and a specific repo load failure are different root causes." + "reason": "Both fail during loading, but one is a tokenizer mapping None case and the other is a physical-intelligence/fast load regression." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different GLM-related bugs with no shared underlying defect." + "reason": "Different subsystems entirely: image processor token counting versus GlmMoeDsaIndexer torch.split handling." }, { "left": "issue:43097", "right": "issue:45588", "accept": false, - "reason": "Config-removal regression is unrelated to the flash_attention.py s_aux=None crash." + "reason": "A config/API removal issue is unrelated to a flash_attention.py crash on sink-less models." }, { "left": "issue:43673", "right": "issue:46123", "accept": false, - "reason": "Both touch generation/pipeline mechanics, but they are separate bugs in different components." + "reason": "Both touch generation-related code, but chunked_prefill cache handling and MaskGenerationPipeline batching are distinct bugs." }, { "left": "issue:45440", "right": "issue:46172", "accept": false, - "reason": "Both are DeepSeek-related, but they describe different models and different errors." + "reason": "Both are DeepSeek-related, but they target different model versions and different concrete defects." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Both are Qwen3 family issues, but one is missing projections and the other is a missing vision-config field." + "reason": "Qwen3next missing projections and Qwen3_5MoeVisionConfig missing a field are separate model/config issues." }, { "left": "issue:44038", "right": "issue:44410", "accept": false, - "reason": "Same model family, but the reported incompatibilities are different and not clearly one bug." + "reason": "Different Qwen-family regressions with different failure modes; the titles do not indicate the same underlying bug." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are ImportErrors, but they involve different missing symbols from different packages." + "reason": "Both are ImportErrors, but they are for different missing symbols from different packages and likely separate dependency/version issues." }, { "left": "issue:43097", "right": "issue:46050", "accept": false, - "reason": "Config removal and quantized save/load breakage are unrelated." + "reason": "Config removal and quantized save/load breakage are unrelated regressions." }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "Different training-stack failures: FSDP2/PEFT optimizer-state corruption vs Liger Kernel not applied with model_init." + "reason": "Different training stack problems: optimizer state corruption under FSDP2/PEFT versus Liger Kernel not being applied with model_init." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Processor loading without torchvision and ReasoningEffort import failure are different dependency/import problems." + "reason": "Both are loading-time crashes, but one is about missing torchvision for video processors and the other is a mistral_common import breakage." }, { "left": "issue:43208", "right": "issue:46897", "accept": false, - "reason": "Different models and different bug classes; no sign of a shared underlying defect." + "reason": "The xLSTM report bundles multiple training bugs, while the Florence2 issue is a specific double-shift loss bug; not the same defect." }, { "left": "issue:46655", "right": "issue:46664", "accept": false, - "reason": "Same SmolVLM video processor area, but the titles point to different bugs (padding calculation vs num_frames semantics) without evidence they are one issue." + "reason": "Both are SmolVLM video-processor issues, but incorrect padding calculation and num_frames semantics are distinct concrete bugs." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different modules; they are similar in shape but not the same bug." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "A flaky integration test from missing set_seed is unrelated to a noisy logging message." + }, + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "No meaningful overlap: log noise versus an IndexError in apply_chat_template on empty input." } ] }, @@ -17402,7 +17409,6 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-517", "nodes": [ - "issue:38617", "issue:42890", "issue:42907", "issue:43012", @@ -17424,7 +17430,7 @@ "issue:44077", "issue:44117", "issue:44162", - "issue:44351", + "issue:44222", "issue:44368", "issue:44380", "issue:44451", @@ -17435,6 +17441,7 @@ "issue:44936", "issue:44938", "issue:45103", + "issue:45478", "issue:45584", "issue:45712", "issue:45753", @@ -17446,15 +17453,12 @@ "issue:46421", "issue:46701", "issue:46710", - "issue:46752", "issue:46821", + "issue:47059", "issue:47116" ], "soft_pairs": [ - "issue:38617|issue:44351", - "issue:42890|issue:46298", "issue:44062|issue:44117", - "issue:46298|issue:46752", "issue:43986|issue:44484", "issue:43097|issue:43873", "issue:43012|issue:46821", @@ -17471,10 +17475,10 @@ "issue:42907|issue:44655", "issue:43012|issue:43550", "issue:45712|issue:46298", - "issue:43012|issue:44368", "issue:43992|issue:44484", - "issue:43208|issue:43550", + "issue:43012|issue:44368", "issue:43827|issue:47116", + "issue:43208|issue:43550", "issue:43352|issue:43976", "issue:43352|issue:44380", "issue:43066|issue:43906", @@ -17484,240 +17488,236 @@ "issue:45584|issue:46365", "issue:44380|issue:46132", "issue:44162|issue:46172", - "issue:44077|issue:46365" + "issue:44077|issue:46365", + "issue:44222|issue:45478", + "issue:42890|issue:47059" ], - "prepared_review_unit_hash": "6a7cbba4bef01b04ee730bbe3a27933f1a9f758914c09c2e0d473312ba0eab43", + "prepared_review_unit_hash": "eb8cd46dba2d5a6d29552c8ec0c944e4a43a4176f89f56c447c71b8d9a659f47", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 34, - "serialized_chars": 37148, - "estimated_input_tokens": 9287, - "estimated_eval_tokens": 18830 + "soft_pair_count": 33, + "serialized_chars": 36820, + "estimated_input_tokens": 9205, + "estimated_eval_tokens": 18666 }, "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 set is a collection of unrelated issue reports across different code paths. No PRs are present.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "The set is mostly heterogeneous, with a loose subtheme of Transformers v5 regressions and compatibility issues, but no soft pair looks like a true duplicate or same underlying fix. I would anchor triage on the broad v5 streaming-regression report, while treating the rest as separate bugs.", + "confidence": 0.43, + "canonical_issue_reason": "Issue 47116 is the clearest broad anchor: it explicitly describes a Transformers v5 runtime regression in the streaming pipeline, which best matches the small subset of v5-related breakage reports here.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43502 is the clearest, most actionable standalone bug report in the set, but it is not a duplicate of the others.", + "best_issue_reason": "47116 is the most suitable global triage anchor because it is broad, recent, and concrete; however the overall set is mostly heterogeneous and not a single duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:38617", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but for different missing symbols and different failure modes; not the same bug." - }, - { - "left": "issue:42890", - "right": "issue:46298", - "accept": false, - "reason": "One is a flaky test seed issue; the other is a noisy logging complaint. Different problems." - }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Different tokenizer bugs: AddedToken constructor misuse vs tokenizer mapping returning None." - }, - { - "left": "issue:46298", - "right": "issue:46752", - "accept": false, - "reason": "Noise/logging versus an apply_chat_template IndexError; unrelated code paths." + "reason": "Different tokenizer-loading failures: AddedToken keyword collision vs TOKENIZER_MAPPING_NAMES returning None." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "AutoProcessor/torchvision loading crash is unrelated to save_pretrained shard size behavior." + "reason": "Unrelated: missing torchvision during AutoProcessor loading vs a save_pretrained shard-size question." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "Config-flag removal complaint and quantization offloading bug are separate issues." + "reason": "Different bugs: a removed config/API in v5 vs quantization offloading behavior." }, { "left": "issue:43012", "right": "issue:46821", "accept": false, - "reason": "PyTorch precision warning in bfloat16 compilation is unrelated to TimesFM window_size AttributeError." + "reason": "Different models and failure modes: bf16 precision warning vs TimesFM window_size AttributeError." }, { "left": "issue:44117", "right": "issue:46701", "accept": false, - "reason": "Tokenizer mapping None handling is unrelated to an image-text-to-text pipeline failure." + "reason": "Tokenizer mapping null-handling vs an image-text pipeline failure are not the same code path." }, { "left": "issue:46172", "right": "issue:46710", "accept": false, - "reason": "A DeepSeek tensor transpose bug is not the same as an accuracy regression after upgrade." + "reason": "Different DeepSeek issues: one is a dtype/transpose bug, the other is an accuracy regression after upgrade." }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "Local-files-only network leakage and missing embed_tokens.weight are different loading bugs." + "reason": "Different loader problems: local_files_only network leak vs missing embed_tokens.weight in UMT5Encoder." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "Passing kwargs to cached_file is unrelated to auto_docstring crashing on future annotations." + "reason": "Unrelated: cached_file kwargs passthrough vs docstring parsing crash with future annotations." }, { "left": "issue:44936", "right": "issue:46365", "accept": false, - "reason": "Trainer evaluate-after-train behavior and Trainer import interference from datasets are different failures." + "reason": "Trainer evaluation after train is a runtime state bug; import failure due to datasets is a dependency/import issue." }, { "left": "issue:44451", "right": "issue:47116", "accept": false, - "reason": "Model loading failure and streaming pipeline regression do not share the same concrete root cause." + "reason": "Different reports: model loading failure for ScandiBERT vs a streaming pipeline regression in v5." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "GPT2 attention scaling with SDPA/FlashAttention is unrelated to chunked prefill cache recompilation." + "reason": "Different subsystems: GPT2 attention scaling under SDPA/FlashAttention vs cache lazy-init recompilation." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Python 3.14 import failure and BACKENDS_MAPPING tensorflow_text validation are separate issues." + "reason": "Python version import failure and tensorflow_text backend mapping are unrelated." }, { "left": "issue:43064", "right": "issue:47116", "accept": false, - "reason": "Distributed optimizer-state corruption and streaming pipeline regression are unrelated." + "reason": "Different code paths: distributed optimizer-state corruption vs streaming pipeline regression." }, { "left": "issue:42907", "right": "issue:45753", "accept": false, - "reason": "Saving dequantized models and an infinite-loop multimodal bug are different code paths." + "reason": "Save/dequantization failure vs an image-generation infinite loop are not the same bug." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Both mention saving, but one is dequantized model save failure and the other is Pipeline.save_pretrained support." + "reason": "Both involve saving, but one is dequantized model saving and the other is Pipeline.save_pretrained; different objects and failure modes." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "Precision warning during bfloat16 compilation is unrelated to Bamba torch.compile SDPA failure." + "reason": "A PyTorch warning about bf16 precision is unrelated to a torch.compile SDPA failure." }, { "left": "issue:45712", "right": "issue:46298", "accept": false, - "reason": "Left is dummy-class cleanup; right is a logging message complaint." - }, - { - "left": "issue:43012", - "right": "issue:44368", - "accept": false, - "reason": "A compile-time warning and a tie_word_embeddings config warning are not the same issue." + "reason": "Left is a repo hygiene/dummy-class issue; right is just a logging complaint." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "Missing encoder weights on load is unrelated to save_pretrained shard size defaults." + "reason": "Missing encoder weights on load is unrelated to the max_shard_size default question." }, { - "left": "issue:43208", - "right": "issue:43550", + "left": "issue:43012", + "right": "issue:44368", "accept": false, - "reason": "xLSTM training bugs and Bamba SDPA compile failure are different model-specific problems." + "reason": "Different warnings with different causes: bf16 precision warning vs tie_word_embeddings config warning." }, { "left": "issue:43827", "right": "issue:47116", "accept": false, - "reason": "Docs still referencing pipeline after removals is unrelated to a streaming regression." + "reason": "Docs still referencing pipeline() is not the same underlying bug as a streaming pipeline regression." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba compile/SDPA failure are unrelated." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "Flash Attention support for Nemotron and Python 3.9/3.10 compatibility are distinct issues." + "reason": "FlashAttention support in a model class is unrelated to Python version compatibility." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Nemotron Flash Attention support and GPT2 attention scaling under SDPA are different model/backend bugs." + "reason": "Both mention attention backends, but one is unsupported FlashAttention 2 and the other is ignored GPT2 scaling config." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Wrong tokenizer decoder type is not the same as an isolated reproduction of another issue." + "reason": "Tokenizer decoder type mismatch is not clearly the same as the isolated reproduction of another issue." }, { "left": "issue:43825", "right": "issue:44655", "accept": false, - "reason": "Pipeline error messaging and Pipeline.save_pretrained behavior are separate concerns." + "reason": "Pipeline task-support messaging and Pipeline.save_pretrained are different problems." }, { "left": "issue:44734", "right": "issue:47116", "accept": false, - "reason": "KV-cache tensor indexing crash in serve is unrelated to streaming pipeline regression." + "reason": "A serving KV-cache indexing crash is not the same as a streaming pipeline regression." }, { "left": "issue:43097", "right": "issue:46298", "accept": false, - "reason": "Removed config flag warning and noisy logging message are different problems." + "reason": "Removed v5 API behavior and a noisy log message are unrelated." }, { "left": "issue:45584", "right": "issue:46365", "accept": false, - "reason": "Whisper empty-transcription generation bug is unrelated to Trainer import issues." + "reason": "Whisper empty-transcription generation bug is unrelated to the Trainer import problem." }, { "left": "issue:44380", "right": "issue:46132", "accept": false, - "reason": "GPT2 backend scaling and AttentionInterface.register behavior change are different code paths." + "reason": "GPT2 backend-specific attention scaling and AttentionInterface.register behavior changes are different issues." }, { "left": "issue:44162", "right": "issue:46172", "accept": false, - "reason": "ESM2 breakage and a DeepSeek tensor operation bug are unrelated." + "reason": "ESM2 breakage is unrelated to the DeepSeek dtype/transpose bug." }, { "left": "issue:44077", "right": "issue:46365", "accept": false, - "reason": "patchtsmixer post_init validation and Trainer import interference are not the same bug." + "reason": "patchtsmixer post_init validation and Trainer import failure are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both involve MoE/save/load, but one is FP8 save_pretrained and the other is a from_pretrained error; not the same concrete bug." + }, + { + "left": "issue:42890", + "right": "issue:47059", + "accept": false, + "reason": "Seed-related test flakiness and a cli completion API break are unrelated." } ] }, @@ -17741,12 +17741,13 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-517", "nodes": [ + "issue:29870", "issue:36010", + "issue:36296", "issue:36683", "issue:38617", "issue:42175", "issue:42757", - "issue:42890", "issue:42994", "issue:43012", "issue:43064", @@ -17755,6 +17756,7 @@ "issue:43352", "issue:43475", "issue:43502", + "issue:43525", "issue:43526", "issue:43582", "issue:43611", @@ -17767,7 +17769,6 @@ "issue:43976", "issue:44038", "issue:44077", - "issue:44222", "issue:44230", "issue:44355", "issue:44380", @@ -17780,7 +17781,6 @@ "issue:44938", "issue:45216", "issue:45276", - "issue:45478", "issue:45479", "issue:46050", "issue:46077", @@ -17791,20 +17791,18 @@ "issue:47110" ], "soft_pairs": [ - "issue:44222|issue:45478", - "issue:42890|issue:47059", "issue:42757|issue:43824", "issue:44380|issue:45479", "issue:36683|issue:43582", "issue:43064|issue:44380", "issue:42175|issue:46365", "issue:43012|issue:44683", - "issue:44483|issue:44514", "issue:38617|issue:43502", - "issue:43618|issue:44380", + "issue:44483|issue:44514", "issue:43756|issue:44485", - "issue:45276|issue:46077", + "issue:43618|issue:44380", "issue:43618|issue:43761", + "issue:45276|issue:46077", "issue:36010|issue:42757", "issue:43208|issue:43526", "issue:44038|issue:44230", @@ -17814,250 +17812,255 @@ "issue:43352|issue:43937", "issue:44625|issue:46077", "issue:43611|issue:44655", - "issue:36683|issue:43475", "issue:44355|issue:46877", + "issue:36683|issue:43475", "issue:44655|issue:46050", "issue:43976|issue:44938", - "issue:44938|issue:47059", "issue:36683|issue:44077", + "issue:44938|issue:47059", "issue:43867|issue:45216", "issue:36683|issue:47110", "issue:46077|issue:46082", - "issue:43618|issue:45216" + "issue:43618|issue:45216", + "issue:36683|issue:43525", + "issue:29870|issue:36296" ], - "prepared_review_unit_hash": "78617966268dd63063e9191c26a54c9a45e1beaccce2d9291f99edfbddb1fed8", + "prepared_review_unit_hash": "df625cd503d269ca5cf19663b6a81ee1f6d238c4b27255404398ee05158eb224", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37356, - "estimated_input_tokens": 9339, - "estimated_eval_tokens": 18934 + "serialized_chars": 37387, + "estimated_input_tokens": 9347, + "estimated_eval_tokens": 18950 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly lexical similarity noise: the issues span unrelated import errors, model-specific regressions, packaging/version problems, and backend/runtime bugs. None of the soft pairs look like the same underlying bug or change.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "This set is mostly a grab-bag of unrelated issues. The only convincing duplicate-looking pair is the two CLIP v5 regression reports; the rest differ in model, subsystem, or failure mode too much to merge safely.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 36683 is the strongest standalone bug report here: it is concrete, open, and highly discussed, with a clear model/config regression and active triage signal.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 36683 is the best overall representative because it is specific, impactful, and has the richest discussion trail among the set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44222", - "right": "issue:45478", - "accept": false, - "reason": "Both mention FP8 and MoE, but one is about saving FP8 MoE checkpoints while the other is a from_pretrained load error for Qwen3.5 MoE; different failure paths." - }, - { - "left": "issue:42890", - "right": "issue:47059", - "accept": false, - "reason": "Unrelated failures: missing set_seed in a SAM HQ integration test versus a CLI breakage with huggingface_hub 1.22.0." - }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but for different missing symbols from different modules; not the same bug." + "reason": "Both are import errors, but they involve different symbols and different dependencies; not the same bug." }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Different problems: GPT2 attention scaling ignored in SDPA/FlashAttention versus degenerate zero loss in sequence classification with num_labels=1." + "reason": "One is about GPT-2 attention scaling under SDPA/FlashAttention, the other is a zero-loss classification regression; unrelated failures." }, { "left": "issue:36683", "right": "issue:43582", "accept": false, - "reason": "Different subsystems entirely: Gemma3Config missing vocab_size versus an Apple Silicon caching_allocator_warmup TypeError." + "reason": "Gemma3 config vocab_size is unrelated to Apple Silicon allocator warmup TypeError." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT2 attention-scaling behavior." + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT-2 attention scaling config." }, { "left": "issue:42175", "right": "issue:46365", "accept": false, - "reason": "Packaging/dependency issues, but not the same one: torch extra missing TensorFlow versus Trainer import blocked by datasets." + "reason": "One is backend packaging for TensorFlow, the other is a Trainer import problem tied to datasets; different root causes." }, { "left": "issue:43012", "right": "issue:44683", "accept": false, - "reason": "One is a warning during bfloat16 compilation; the other is a functional failure compiling flex_attention on torch >= 2.9." - }, - { - "left": "issue:44483", - "right": "issue:44514", - "accept": false, - "reason": "Different API paths: /v1/chat/completions request handling versus Qwen2_5_VLProcessor.apply_chat_template batching." + "reason": "PyTorch precision warning during compile and flex_attention compile failure are different issues." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "Distinct issues: missing layer_type_validation import versus local_files_only still triggering network requests." + "reason": "Importing a removed utility and ignoring local_files_only are separate API/behavior bugs." }, { - "left": "issue:43618", - "right": "issue:44380", + "left": "issue:44483", + "right": "issue:44514", "accept": false, - "reason": "CLIPOutput attentions assignment regression is unrelated to GPT2 attention scaling under SDPA/FlashAttention." + "reason": "A chat/completions API rejection and a processor batching crash are different code paths." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Both mention RoPE, but one is about Smollm3 dropping layers and the other is a GLM-5 RoPE implementation discussion." + "reason": "Both mention RoPE, but they concern different models and different bugs, not one shared defect." }, { - "left": "issue:45276", - "right": "issue:46077", + "left": "issue:43618", + "right": "issue:44380", "accept": false, - "reason": "Same model family but different bugs: token embedding resize propagation versus bidirectional attention mask construction." + "reason": "CLIP output regression and GPT-2 attention scaling are unrelated." }, { "left": "issue:43618", "right": "issue:43761", + "accept": true, + "reason": "Both describe the same CLIP v5 forward/output regression: optional outputs from CLIPVisionModel are not being populated correctly." + }, + { + "left": "issue:45276", + "right": "issue:46077", "accept": false, - "reason": "Both concern CLIP, but one is missing attentions in CLIPOutput and the other is hidden_states=None in CLIPVisionModel.forward." + "reason": "Both are Gemma4, but one is embedding resize propagation and the other is bidirectional-attention masking; different bugs." }, { "left": "issue:36010", "right": "issue:42757", "accept": false, - "reason": "Different missing imports from different packages; not the same underlying compatibility break." + "reason": "Different missing imports from different packages; not the same underlying failure." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "Completely different areas: xLSTM training bugs versus BeitImageProcessorFast reduce_labels behavior." + "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated." }, { "left": "issue:44038", "right": "issue:44230", "accept": false, - "reason": "Related model family, but one is a loading bug for Qwen3-VL-Moe and the other is fp8 support/inference work; not the same concrete bug." + "reason": "Both mention Qwen3-VL-Moe, but one is a load/regression issue and the other is fp8 inference support; not mergeable as one fix." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Both involve save_pretrained, but one is quantized model saving and the other is Pipeline object serialization." + "reason": "Quantized model save corruption and Pipeline save_pretrained support are different save paths with different semantics." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Different loading failures: Qwen3VL-Moe regression versus a state_dict sorting-related load error." + "reason": "Both are model-loading problems, but the reported failure modes and triggers differ." }, { "left": "issue:44625", "right": "issue:46365", "accept": false, - "reason": "Unrelated issues: num_labels propagation in Qwen3.5 versus inability to import Trainer until datasets is removed." + "reason": "Qwen3.5 label propagation and Trainer import dependency conflicts are unrelated." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Different model/config problems: Flash Attention 2 unsupported for NemotronHForCausalLM versus invalid GenerationConfig for GLM-5." + "reason": "Flash Attention support for Nemotron and GLM-5 GenerationConfig validation are different model/config issues." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "Different bugs in different models: Qwen3.5 label propagation versus Gemma4 bidirectional attention masking." + "reason": "Label propagation and attention-mask construction are unrelated behaviors." }, { "left": "issue:43611", "right": "issue:44655", "accept": false, - "reason": "Model loading with base_model_prefix is unrelated to saving Pipeline objects." + "reason": "Base-model-prefix loading regression and Pipeline save_pretrained are different mechanisms." }, { - "left": "issue:36683", - "right": "issue:43475", + "left": "issue:44355", + "right": "issue:46877", "accept": false, - "reason": "Different missing-attribute failures in different models: Gemma3Config vocab_size versus Sam3VisionEncoderOutput fpn_position_embeddings." + "reason": "Compiled Python file loading errors and auto_docstring import-time stdout diagnostics are different issues." }, { - "left": "issue:44355", - "right": "issue:46877", + "left": "issue:36683", + "right": "issue:43475", "accept": false, - "reason": "Both are import/runtime annoyances, but one is compiled Python file execution errors and the other is auto_docstring printing lint diagnostics at import time." + "reason": "Gemma3 config vocab_size and SAM3 video encoder missing attribute are distinct model-specific bugs." }, { "left": "issue:44655", "right": "issue:46050", "accept": false, - "reason": "Pipeline save_pretrained serialization is not the same as quantized sinq model save/load corruption." + "reason": "General Pipeline serialization and sinq quantized model persistence are different save/load problems." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Both are Python-version compatibility reports, but they target different interpreter versions and failure modes." + "reason": "Python-version compatibility complaints, but one is about Transformers 5.1 support and the other about Python 3.14 loading; not the same issue." }, { - "left": "issue:44938", - "right": "issue:47059", + "left": "issue:36683", + "right": "issue:44077", "accept": false, - "reason": "Python 3.14 import/load failure is unrelated to the huggingface_hub CLI completion attribute regression." + "reason": "Gemma3 vocab_size and patchtsmixer post_init validation are unrelated." }, { - "left": "issue:36683", - "right": "issue:44077", + "left": "issue:44938", + "right": "issue:47059", "accept": false, - "reason": "Gemma3Config missing vocab_size and patchtsmixer post_init validation are unrelated bugs." + "reason": "Python 3.14 import/load failure and huggingface_hub CLI breakage are different compatibility regressions." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Both are save/load regressions, but one is about state_dict ordering during load and the other is an incorrect Qwen3.5 checkpoint saved by save_pretrained." + "reason": "Different checkpoint/model-loading failures with different triggers and symptoms." }, { "left": "issue:36683", "right": "issue:47110", "accept": false, - "reason": "Different attribute errors in different code paths: Gemma3Config vocab_size versus special_tokens being a list instead of a mapping." + "reason": "Gemma3 vocab_size and special_tokens list/dict handling are unrelated." }, { "left": "issue:46077", "right": "issue:46082", "accept": false, - "reason": "Different configuration validations in different models: Gemma4 attention masking versus LlamaConfig head_dim divisibility." + "reason": "One is Gemma4 attention masking, the other is LlamaConfig head_dim validation; different code paths." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "CLIP output-attentions regression is unrelated to Qwen3.5 save_pretrained checkpoint corruption." + "reason": "CLIP output regression and Qwen3.5 save_pretrained checkpoint corruption are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3 config vocab_size and Llama4 pad_token_id are different config attribute issues." + }, + { + "left": "issue:29870", + "right": "issue:36296", + "accept": false, + "reason": "BART decoder_attention_mask assignment and tensor-parallel training are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays conservative. The chosen canonical issue is reasonably supported by the highest discussion activity, and the one accepted soft pair (43618/43761) is a plausible same-regression CLIP output bug. I did not see any soft-edge verdicts that overstate duplication beyond the available evidence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -18078,16 +18081,15 @@ "issue:29870", "issue:33453", "issue:36296", - "issue:36683", "issue:41669", "issue:42890", "issue:42898", "issue:43010", "issue:43064", "issue:43502", - "issue:43525", "issue:43575", "issue:43618", + "issue:43723", "issue:43867", "issue:43976", "issue:43992", @@ -18100,7 +18102,9 @@ "issue:44355", "issue:44360", "issue:44393", + "issue:44485", "issue:44492", + "issue:44556", "issue:44734", "issue:44855", "issue:44908", @@ -18123,8 +18127,6 @@ "issue:47000" ], "soft_pairs": [ - "issue:36683|issue:43525", - "issue:29870|issue:36296", "issue:43064|issue:44492", "issue:43010|issue:45278", "issue:44075|issue:46365", @@ -18136,8 +18138,8 @@ "issue:44355|issue:44855", "issue:36296|issue:43575", "issue:43010|issue:45850", - "issue:41669|issue:44246", "issue:43867|issue:45375", + "issue:41669|issue:44246", "issue:42890|issue:45561", "issue:43618|issue:44077", "issue:43010|issue:43976", @@ -18151,217 +18153,222 @@ "issue:42890|issue:44112", "issue:42898|issue:46664", "issue:44222|issue:45125", - "issue:44734|issue:46617" + "issue:44734|issue:46617", + "issue:43723|issue:44556", + "issue:43502|issue:44485" ], - "prepared_review_unit_hash": "98d48972575297e7fb27c3c461203b31942b27d901291550c207b0397e906fc3", + "prepared_review_unit_hash": "e87c79f4c3c8db7638d1409efb68c5c8f9a0657e9c3d5e28047500d37bb2a969", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35451, - "estimated_input_tokens": 8863, - "estimated_eval_tokens": 17982 + "serialized_chars": 36006, + "estimated_input_tokens": 9002, + "estimated_eval_tokens": 18260 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is largely heterogeneous; most soft pairs are clearly different bugs or feature requests. The only near-match is the import-time performance pair, but the evidence is still too broad to safely deduplicate as the same underlying change.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 41669 is the most concrete, root-cause-oriented report in the set: it identifies a specific import-time regression source (`import *` in models) and ties it to measurable slowdown. That makes it the best representative if the cluster is meant to capture import performance problems.", + "summary": "Most items are unrelated issues around v5 regressions, model loading, tensor parallelism, and test failures. Only one pair looks plausibly like the same duplicate import-performance bug.", + "confidence": 0.48, + "canonical_issue_reason": "Issue 41669 is the clearest, most actionable report and the only one that plausibly subsumes another item: an import-time regression tied to `import *` usage in model modules.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 41669 is the strongest standalone issue to keep: it is specific, actionable, and describes a plausible fixable cause rather than a vague symptom report.", + "best_issue_reason": "41669 is the strongest representative because it points to a concrete cause and user-visible slowdown; the rest are largely separate bugs in different code paths.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:36683", - "right": "issue:43525", - "accept": false, - "reason": "Both are config-attribute errors, but they affect different models and different missing fields (`vocab_size` vs `pad_token_id`), so they are not the same bug." - }, - { - "left": "issue:29870", - "right": "issue:36296", - "accept": false, - "reason": "Bart generation `decoder_attention_mask` UnboundLocalError is unrelated to tensor-parallel training." - }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "Wrong optimizer state/param handling in FSDP2+PEFT is unrelated to a typo in cache strategy docs/code." + "reason": "Wrong subsystem and symptom: optimizer-state/PEFT training bug vs a cache-strategy typo." }, { "left": "issue:43010", "right": "issue:45278", "accept": false, - "reason": "`@torch.no_grad` decoration for cache updates is a specific behavior fix; `Many import errors after update` is a broad upgrade complaint with no same code path." + "reason": "Both mention v5 regressions, but one is about cache update semantics and the other about import failures; not the same bug." }, { "left": "issue:44075", "right": "issue:46365", "accept": false, - "reason": "SGD argument handling and failing to import Trainer with datasets installed are different failures in different code paths." + "reason": "Unrelated: SGD args not used vs importing Trainer fails when datasets is installed." }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "Missing `embed_tokens.weight` for UMT5Encoder and Qwen3.5 MoE `from_pretrained` errors are distinct model-loading problems." + "reason": "Different model families and different failures; no clear shared code-path." }, { "left": "issue:44908", "right": "issue:46851", "accept": false, - "reason": "Scheduler kwargs being ignored is unrelated to Qwen3.5 variable-length position-id handling." + "reason": "Scheduler kwargs bug vs variable-length position_ids bug; unrelated." }, { "left": "issue:29870", "right": "issue:33453", "accept": false, - "reason": "Bart `generate()` mask initialization and tokenizer-loading regression are not the same defect." + "reason": "Bart generate() decoder mask error vs tokenizer-loading regression; different code paths." }, { "left": "issue:44393", "right": "issue:47000", "accept": false, - "reason": "Qwen3-VL bounding-box hallucination and RTDetrLoss infeasible cost matrix are different vision-model failures." + "reason": "Different models and failures: Qwen3-VL bbox output vs RTDetr loss infeasible cost matrix." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "State-dict sorting load errors and Gemma pipeline roundtrip on `\u2581` are not the same underlying issue." + "reason": "Model state_dict loading error vs Gemma pipeline roundtrip/tokenization issue; not the same bug." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Compiled Python-file import errors and a Python 3.13 `@torch.jit.script` parse/IndentationError are different failure modes." + "reason": "Both are import-related, but one is about compiled Python files and the other is a torch.jit parsing/IndentationError on Python 3.13." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Tensor-parallel training bug and OOM when loading Qwen2-57B with TP are related only at a high level; they do not look like the same concrete bug." + "reason": "Both involve tensor parallelism, but one is a training bug and the other is an OOM while loading a specific Qwen model." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "A specific cache-update `no_grad` fix is not the same as a vague claim that v5 breaks too many models." - }, - { - "left": "issue:41669", - "right": "issue:44246", - "accept": false, - "reason": "Both concern import slowness, but one is a concrete root-cause proposal (`import *` in models) and the other is a vague symptom report; not enough evidence they are the same exact bug/change." + "reason": "One is a specific cache/no_grad API change; the other is a broad complaint about many models breaking." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "State-dict loading sortedness and missing `deepstack_visual_indexes` in Qwen3.5 vision config are different issues." + "reason": "State-dict loading failure vs missing vision-config field; different underlying problems." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": true, + "reason": "Both are about `transformers` import taking too long; 41669 identifies the likely root cause (`import *` in model modules), and 44246 reports the same slowdown symptom." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "A missing `set_seed()` in SAM HQ tests is unrelated to a pytest-xdist race on `captured_info.txt`." + "reason": "Test flakiness from missing seeds vs xdist race on a captured-info file; not the same issue." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIPOutput attentions assignment and patchtsmixer `post_init` allowance are unrelated model-specific changes." + "reason": "Different model/API bugs: missing attentions assignment vs optional `post_init` deprecation/constraint." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "Cache `update()` decoration and Python-version compatibility are unrelated." + "reason": "Cache update no_grad bug vs scheduler argument handling; unrelated." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "UMT5Encoder missing weights and Qwen3.5 MoE `from_pretrained` regression are different model-loading problems." + "reason": "Same broad area of model loading, but different model versions and no concrete shared failure mode." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Generic load error with sorted state_dict and incorrect Qwen3.5 visual encoder keys are not the same bug." + "reason": "State_dict loading error vs incorrect visual encoder keys on save_pretrained; different directions and symptoms." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "Missing PyTorch dependency import error is unrelated to tokenizer-loading regression." + "reason": "PyTorch missing ImportError vs tokenizer-loading regression; unrelated." }, { "left": "issue:43618", "right": "issue:46285", "accept": false, - "reason": "CLIP attentions assignment and CLIP text-model structure inconsistency are different CLIP-related changes, not the same defect." + "reason": "Different CLIP-related structure regressions; one is about attentions, the other about `.text_model` layout divergence." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "A missing ReLU in the DSA indexer is unrelated to DeepSeekV3 MoE divergence from remote implementation." + "reason": "DSA indexer discussion vs DeepSeekV3 implementation divergence; no shared concrete bug." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "Test seeding flakiness is unrelated to inability to load Phi-4 multimodal instruct." + "reason": "Seed-related test instability vs model loading failure for Phi-4 multimodal; unrelated." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "`local_files_only=True` network requests and `AddedToken(special=...)` argument duplication are different code paths." + "reason": "Local-files-only API leakage vs AddedToken constructor error; different code paths." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "SAM HQ test seeding flakes and GraniteSpeech stale device override CI failure are unrelated test issues." + "reason": "Seed-related test flakiness vs stale device override test on CI; different issues." }, { "left": "issue:42898", "right": "issue:46664", "accept": false, - "reason": "Tokenizer cleanup behavior changes and SmolVLM video-frame semantics are unrelated." + "reason": "Tokenizer cleanup behavior change vs video processor `num_frames` semantics; unrelated." }, { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "FP8 `save_pretrained` for MoE and missing `_tp_plan` for tensor parallelism are different functionality gaps." + "reason": "FP8 save_pretrained issue vs missing tensor-parallel plan for Qwen3_5MoE; different problems." }, { "left": "issue:44734", "right": "issue:46617", "accept": false, - "reason": "KV-cache continuation tensor indexing crash and trust-remote-code cache `FileNotFoundError` are not the same failure." + "reason": "KV-cache continuation crash in serve vs trust_remote_code cache FileNotFoundError; unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer.from_pretrained regression vs checkpoint reload incompatibility after upgrade; related area, but not the same concrete bug." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Local-files-only network leak vs RoPE implementation discussion; unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted soft-edge pair, issue:41669 with issue:44246, is well-supported by their shared import-slowdown symptom, and the rejections do not overclaim equivalence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -18397,12 +18404,12 @@ "issue:43475", "issue:43502", "issue:43576", - "issue:43723", "issue:43824", "issue:43827", "issue:43874", "issue:43901", "issue:44222", + "issue:44230", "issue:44315", "issue:44351", "issue:44355", @@ -18428,12 +18435,10 @@ "issue:47110" ], "soft_pairs": [ - "issue:43723|issue:44556", - "issue:43502|issue:44485", "issue:43874|issue:46617", - "issue:43010|issue:43065", "issue:43824|issue:46617", "issue:45850|issue:47110", + "issue:43010|issue:43065", "issue:43576|issue:46132", "issue:42548|issue:44351", "issue:45125|issue:45478", @@ -18466,264 +18471,235 @@ "issue:43475|issue:46617", "issue:46143|issue:46144", "issue:41628|issue:42548", - "issue:45375|issue:47110" + "issue:45375|issue:47110", + "issue:44222|issue:44230" ], - "prepared_review_unit_hash": "450c5cb863c2adc33d2dbfc5578a9b8f1ec15a1e652012a7a20e0cd818e8671c", + "prepared_review_unit_hash": "1ac602f56978a792778915b3dab834bcd6075c8d4571750f1e7be1e892bc6c59", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 39, - "serialized_chars": 38909, - "estimated_input_tokens": 9728, - "estimated_eval_tokens": 19712 + "soft_pair_count": 38, + "serialized_chars": 38592, + "estimated_input_tokens": 9648, + "estimated_eval_tokens": 19552 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of Transformers v5 regressions, loading/import failures, and unrelated subsystem bugs. None of the soft pairs are convincing duplicates on the evidence shown, so I would not merge any of them as the same underlying issue.", - "confidence": 0.86, - "canonical_issue_reason": "issue:44556 is the broadest concrete regression report here: it directly describes v5 upgrade breakage when reloading a checkpoint, which is the closest umbrella theme shared by several other loading-related reports.", + "summary": "This set is heterogeneous: most items are unrelated import, loading, CLI, docs, quantization, or model-specific regressions. I found no true duplicate pair among the soft edges, so the cluster should not be merged into a single canonical bug/change.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44556 is the best representative because it is concrete, version-specific, and centrally about the v5 loading regression theme rather than a narrow model or API detail.", + "best_issue_reason": "No single issue safely represents the whole set; issue:43576 is the broadest/highest-signal regression, but it still does not subsume the other unrelated reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43723", - "right": "issue:44556", - "accept": false, - "reason": "Both involve loading regressions after v5, but one is tokenizer loading and the other is checkpoint/model reload; not the same code-path failure." - }, - { - "left": "issue:43502", - "right": "issue:44485", - "accept": false, - "reason": "Unrelated bugs: local_files_only network behavior versus a GLM-5 RoPE implementation discussion." - }, { "left": "issue:43874", "right": "issue:46617", "accept": false, - "reason": "Different failures: a missing image-processor method in multimodal token counting versus cache loading FileNotFoundError for trust_remote_code." - }, - { - "left": "issue:43010", - "right": "issue:43065", - "accept": false, - "reason": "Different components and symptoms: cache/layer update no_grad semantics versus a dummy Conv2d in Sam3PixelDecoder." + "reason": "Different failures: GLM46V multimodal token counting vs trust_remote_code cache loading FileNotFoundError." }, { "left": "issue:43824", "right": "issue:46617", "accept": false, - "reason": "Import error for a specific Qwen2.5 VL class is not the same as a cached trust_remote_code FileNotFoundError." + "reason": "Different import problem and model family; no shared concrete code path." }, { "left": "issue:45850", "right": "issue:47110", "accept": false, - "reason": "Too generic versus a specific special_tokens attribute error; no shared concrete bug." + "reason": "Broad complaint about many v5 breakages vs a specific special_tokens structure error." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "Cache update no_grad concern vs Sam3PixelDecoder dummy Conv2d; unrelated fixes." }, { "left": "issue:43576", "right": "issue:46132", "accept": false, - "reason": "One is the v5 env CLI command, the other is AttentionInterface.register behavior; different code paths." + "reason": "CLI env command regression vs AttentionInterface.register behavior change; different bugs." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different symbols and likely unrelated causes." + "reason": "Different missing exports: PreTrainedModel vs HybridCache." }, { "left": "issue:45125", "right": "issue:45478", "accept": false, - "reason": "Same model family, but one is missing _tp_plan and the other is a from_pretrained error; not enough to treat as the same defect." + "reason": "Qwen3.5 MoE tensor-parallel plan issue vs from_pretrained loading error; not the same concrete bug." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "One is int4 quantization memory/OOM behavior, the other is checkpoint reload failure after version upgrade." + "reason": "Int4 quantization OOM vs checkpoint reload incompatibility after version upgrade." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLFlow job-reporting failure and tokenizer-loading regression are unrelated." + "reason": "MLFlow job reporting bug vs tokenizer loading regression." }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled Python file errors and slower filtering after upgrade are different issues." + "reason": "Compiled Python file errors vs performance slowdown after upgrade; unrelated." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Both touch Qwen3.5 MoE, but one is tensor-parallel planning and the other is a missing config field; different bugs." + "reason": "Tensor-parallel plan missing vs config field dropped by strict validation; different model issues." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Sentence Transformers loading trouble is unrelated to forward() doc confusion." + "reason": "Loading hang vs confusing forward-argument docs." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Documentation about pipeline removals and a meaningless 'racoon' issue do not match." + "reason": "Docs still reference pipeline() vs unrelated 'racoon' issue." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "No overlap beyond being issues in the repo; different topics entirely." + "reason": "local_files_only network leak vs unrelated 'racoon' issue." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "local_files_only network requests and FP8 save_pretrained moe are unrelated." + "reason": "Offline-mode network leak vs FP8 save_pretrained MoE bug." }, { "left": "issue:46664", "right": "issue:46726", "accept": false, - "reason": "Both are processor-related, but one is num_frames semantics and the other is missing vision backend handling." + "reason": "Different video-processor semantics vs missing-vision-backend AutoProcessor error." }, { "left": "issue:43010", "right": "issue:43232", "accept": false, - "reason": "Cache/layer update no_grad semantics are not the same as generation kwargs sync_gpus ordering." + "reason": "Cache update decoration issue vs generation kwargs after sync_gpus." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLFlow reporting on failed jobs is unrelated to tensor-parallel training behavior." + "reason": "MLFlow reporting failure vs tensor parallel training bug." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Both mention version-related loading breakage, but for different model families and different failure modes." + "reason": "Qwen3VL MoE load regression vs generic checkpoint reload incompatibility." }, { "left": "issue:29870", "right": "issue:31795", "accept": false, - "reason": "Bart generate() UnboundLocalError is unrelated to model.forward documentation wording." + "reason": "BART generate() UnboundLocalError vs docs confusion about forward inputs." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Two different loading regressions affecting different artifacts/models; not the same defect." + "reason": "Tokenizer loading regression vs Llama 3.2 Vision model-loading failure; different affected paths." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled-file runtime errors and PretrainedConfig type checking are separate problems." + "reason": "Compiled-file runtime errors vs PretrainedConfig type-checking regression." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Different breakages: huggingface_hub import error versus the transformers env CLI command." + "reason": "Missing huggingface_hub symbol vs transformers env CLI regression." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "CLI command breakage and PretrainedConfig type-checking are unrelated." + "reason": "CLI env failure vs type-checking regression; unrelated." }, { "left": "issue:43576", "right": "issue:47059", "accept": false, - "reason": "Both are CLI-adjacent, but one is the env command and the other is a huggingface_hub compatibility issue in tests/cli." + "reason": "CLI env command regression vs tests/cli hub completion attribute error; not the same concrete bug." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Documentation confusion and an import error from configuration_utils are not the same issue." + "reason": "Docs confusion vs missing layer_type_validation import error." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "No relation between a placeholder issue and the Gemma-4 audio positional encoding bug." + "reason": "Unrelated 'racoon' placeholder issue vs Gemma-4 audio positional encoding bug." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "CLI env-command failure and a missing Qwen2.5 VL class import are unrelated." + "reason": "CLI env regression vs missing Qwen2_5_VL export; different issues." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger Kernel application in model_init and a DSA indexer ReLU discussion are unrelated." - }, - { - "left": "issue:44485", - "right": "issue:44961", - "accept": false, - "reason": "GLM-5 RoPE discussion and the placeholder issue are unrelated." + "reason": "Liger kernel application bug vs DSA indexer activation discussion; unrelated." }, { - "left": "issue:44351", - "right": "issue:46617", + "left": "issue:41628", + "right": "issue:42548", "accept": false, - "reason": "HybridCache import error and cache FileNotFoundError are different failure modes." + "reason": "Different missing imports: AutoImageProcessor vs PreTrainedModel." }, { - "left": "issue:43352", - "right": "issue:44961", + "left": "issue:45375", + "right": "issue:47110", "accept": false, - "reason": "Flash Attention support error for Nemotron and the placeholder issue are unrelated." + "reason": "Missing config field vs special_tokens list shape error; unrelated." }, { - "left": "issue:43475", - "right": "issue:46617", + "left": "issue:44222", + "right": "issue:44230", "accept": false, - "reason": "SAM3 VisionEncoderOutput attribute error is not the same as loading a remote-code model from cache." + "reason": "FP8 save_pretrained bug vs FP8 inference support request; same broad area, different concrete change." }, { "left": "issue:46143", "right": "issue:46144", "accept": false, - "reason": "Related RoFormer area, but one is kwargs forwarding and the other is attention-interface usage; not clearly the same concrete bug." - }, - { - "left": "issue:41628", - "right": "issue:42548", - "accept": false, - "reason": "Different missing imports: AutoImageProcessor versus PreTrainedModel." - }, - { - "left": "issue:45375", - "right": "issue:47110", - "accept": false, - "reason": "Missing config field silently dropped and special_tokens list attribute error are unrelated." + "reason": "kwargs propagation bug vs attention-interface usage bug in RoFormer; related theme but not the same underlying defect." } ] }, @@ -18760,7 +18736,6 @@ "issue:43976", "issue:44038", "issue:44077", - "issue:44222", "issue:44230", "issue:44246", "issue:44273", @@ -18797,13 +18772,12 @@ "issue:47176" ], "soft_pairs": [ - "issue:44222|issue:44230", "issue:44315|issue:46209", "issue:43867|issue:47110", "issue:43525|issue:47110", - "issue:43673|issue:44492", "issue:44485|issue:46664", "issue:45362|issue:45850", + "issue:43673|issue:44492", "issue:44623|issue:45230", "issue:44273|issue:45542", "issue:44038|issue:45125", @@ -18815,8 +18789,8 @@ "issue:44393|issue:45753", "issue:43525|issue:45850", "issue:44230|issue:46095", - "issue:44556|issue:45092", "issue:43572|issue:45850", + "issue:44556|issue:45092", "issue:44492|issue:45850", "issue:42371|issue:44492", "issue:45850|issue:47059", @@ -18838,280 +18812,268 @@ "issue:43976|issue:46218", "issue:44077|issue:44861" ], - "prepared_review_unit_hash": "f00b83575dc9aefe34d78a2a7ee7fffc81271a3e409a19542c3af5e8331ba305", + "prepared_review_unit_hash": "58727041a14257a430335955bc40ed70af2a6c90f59115054eb41eb33b0ea17b", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 40, - "serialized_chars": 39004, - "estimated_input_tokens": 9751, - "estimated_eval_tokens": 19758 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 39, + "serialized_chars": 38175, + "estimated_input_tokens": 9544, + "estimated_eval_tokens": 19344 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is loosely centered on Transformers v5 regressions and model/config import issues, but the soft pairs mostly overlap only at the subsystem or version level rather than the same concrete bug. I did not accept any soft-edge duplicates.", - "confidence": 0.61, - "canonical_issue_reason": "issue:45850 is the best umbrella representative: it is the broadest, most central v5-regression report and matches the general theme of several other issues without being tied to one model or code path.", + "summary": "This cluster is mostly a loose collection of unrelated transformer/model bug reports that only share superficial API/model-family similarity. I found no soft pair that clearly represents the same underlying bug or change, so all soft edges should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "No single strong canonical duplicate target stands out; issue #44273 is the closest umbrella candidate because it is open and broadly about lazy-loading/import behavior, but the overall cluster is too heterogeneous for a reliable canonical issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45850 is the strongest anchor for this cluster because it is generic, broad, and likely to absorb the nearby v5-breakage reports better than the more specific model/config issues.", + "best_issue_reason": "#44273 is the best available issue to keep as a representative anchor only because it is open and relatively broad; however, it should not be treated as a true duplicate hub for the rest of this set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44222", - "right": "issue:44230", - "accept": false, - "reason": "Both mention FP8/MoE, but one is save_pretrained behavior and the other is inference support; different failure modes and code paths." - }, { "left": "issue:44315", "right": "issue:46209", "accept": false, - "reason": "Unrelated topics: Liger Kernel/model_init application versus FP8 inference on B200." + "reason": "Different problems: model_init/Liger-Kernel application versus FP8 inference failure on B200." }, { "left": "issue:43867", "right": "issue:47110", "accept": false, - "reason": "One is state_dict loading order, the other is special_tokens structure; no shared concrete bug." + "reason": "One is a state_dict loading-order error; the other is a special_tokens type error. No shared code-path is evident." }, { "left": "issue:43525", "right": "issue:47110", "accept": false, - "reason": "Pad-token config attribute error is unrelated to a special_tokens list/dict mismatch." - }, - { - "left": "issue:43673", - "right": "issue:44492", - "accept": false, - "reason": "Cache missing during chunked prefill is not the same as a typo in cache strategy text/metadata." + "reason": "Llama4Config pad_token_id missing is unrelated to special_tokens list/dict handling." }, { "left": "issue:44485", "right": "issue:46664", "accept": false, - "reason": "RoPE implementation discussion versus video processor num_frames semantics; different components and bugs." + "reason": "RoPE implementation for GLM-5 and video num_frames semantics are unrelated subsystems." }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "45362 is a specific chat crash for Qwen3.5-35B; 45850 is a generic v5 breakage umbrella." + "reason": "Specific chat crash on Qwen3.5-35B is not the same underlying bug as a vague 'many models break in v5' complaint." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression and a cache-strategy typo are different issues." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "processor.save_pretrained missing files is a concrete serialization bug; 45230 is too generic to be the same issue." + "reason": "Missing processor save_pretrained files is not the same as an uninformative generic bug report." }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Lazy loading/import behavior and missing TensorFlow backend handling are different problems." + "reason": "Lazy loading issues and TensorFlow backend detection are distinct failure modes." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "Qwen3-VL-Moe bug report versus missing tensor-parallel plan; not the same underlying defect." + "reason": "Qwen3-VL-Moe bug versus missing tensor-parallel plan for Qwen3_5Moe; different concrete defects." }, { "left": "issue:44360", "right": "issue:46664", "accept": false, - "reason": "DSA indexer activation issue is unrelated to a video processor num_frames inconsistency." + "reason": "DSA indexer activation discussion is unrelated to SmolVLM frame-count semantics." }, { "left": "issue:44273", "right": "issue:45003", "accept": false, - "reason": "Both relate broadly to loading/import internals, but one is lazy loading behavior and the other is unsafe sys.modules access; not a duplicate." + "reason": "These are both import-related, but one is a generic lazy-loading complaint and the other is a specific sys.modules access issue; not enough to confirm same bug." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Import-star performance regression is unrelated to a generation cache bug." + "reason": "Import-time performance regression from import * usage is not the same as a chunked_prefill cache bug." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "Qwen3-VL bounding box error and PretrainedConfig type-checking regression are unrelated." + "reason": "Qwen3-VL bbox hallucination/error and PretrainedConfig type-checking breakage are unrelated." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "Both are v5 regressions, but one is the env CLI and the other is checkpoint reload compatibility; different code paths." + "reason": "Broken env command and checkpoint reload incompatibility are different versioning issues." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Both touch Qwen3-VL-family image behavior, but one is 2D box output and the other is an infinite loop on a specific image; too different to merge." + "reason": "Both involve Qwen3_5/VL, but one is bbox output corruption and the other is an infinite-loop image case; not the same bug." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "A missing pad_token_id attribute in one config is a specific model config bug, not the broad v5 breakage report." + "reason": "A missing config attribute is not the same as a broad report that many models break in v5." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 inference support and a weight-conversion-script question are not the same bug." + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5-Moe does not match a deepseekv4 weight-conversion question." }, { - "left": "issue:44556", - "right": "issue:45092", + "left": "issue:43572", + "right": "issue:45850", "accept": false, - "reason": "Both are upgrade/loading incompatibilities, but the affected mechanisms differ: checkpoint reload versus remote-code meta initialization." + "reason": "StableLm pad_token_idx regression is a specific config issue, not the generic v5 breakage report." }, { - "left": "issue:43572", - "right": "issue:45850", + "left": "issue:44556", + "right": "issue:45092", "accept": false, - "reason": "StableLmConfig pad_token_idx regression is specific; 45850 is only a broad umbrella." + "reason": "Both are v5 compatibility complaints, but one is checkpoint reload compatibility and the other is remote-code meta-init incompatibility; different defects." }, { "left": "issue:44492", "right": "issue:45850", "accept": false, - "reason": "A typo in cache-strategy docs/code is not the same as broad model breakage after v5." + "reason": "A cache-strategy typo is not the same as a broad 'many models break' report." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API settings versus cache-strategy typo are unrelated." + "reason": "TF32 API guidance and cache-strategy naming are unrelated." }, { "left": "issue:45850", "right": "issue:47059", "accept": false, - "reason": "General v5 breakage and a specific CLI incompatibility with huggingface_hub are not the same issue." + "reason": "General model breakage in v5 is unrelated to CLI completion errors caused by huggingface_hub changes." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import performance from import * usage is unrelated to a specific module import failure for Qwen3_5ForCausalLM." + "reason": "Import-time slowdown from import * does not match a module import failure for Qwen3_5ForCausalLM." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "sam3_video/sam3_tracker type warning versus Liger Kernel application issue are unrelated." - }, - { - "left": "issue:44295", - "right": "issue:45230", - "accept": false, - "reason": "Buffer access error for position_ids is a concrete tensor/buffer bug; 45230 is too generic." + "reason": "Model-type mismatch warning and Liger-Kernel/model_init behavior are unrelated." }, { - "left": "issue:44297", - "right": "issue:45230", + "left": "issue:44246", + "right": "issue:44273", "accept": false, - "reason": "Tokenizer class mismatch on save_pretrained is a specific serialization bug, not a generic report." + "reason": "Import slowness and lazy loading malfunction are adjacent but not the same concrete bug based on the titles alone." }, { - "left": "issue:43976", - "right": "issue:44393", + "left": "issue:42890", + "right": "issue:44908", "accept": false, - "reason": "Python version compatibility regression and Qwen3-VL bbox output error are unrelated." + "reason": "Missing set_seed in tests is unrelated to inverse_sqrt scheduler kwargs handling." }, { - "left": "issue:44908", - "right": "issue:45071", + "left": "issue:43576", + "right": "issue:44908", "accept": false, - "reason": "Scheduler kwargs handling and PretrainedConfig type-checking are different areas and failure modes." + "reason": "Broken env command and scheduler kwargs omission are unrelated." }, { - "left": "issue:44492", + "left": "issue:43576", "right": "issue:47176", "accept": false, - "reason": "Cache-strategy typo is unrelated to a shell-operator command-substitution allowlist bypass." + "reason": "Environment command breakage is unrelated to a shell-command allowlist bypass security bug." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Nemotron flash-attention support error is unrelated to Qwen3-VL bounding-box hallucination." + "reason": "Flash Attention 2.0 support limitation for Nemotron and Qwen3-VL bbox issues are different model bugs." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "Position_ids buffer access bug does not match a generic placeholder issue." + "reason": "position_ids buffer handling is unrelated to the generic bug report." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer save_pretrained mismatch is a concrete bug; 45230 lacks specificity and is not the same defect." + "reason": "Tokenizer_class mismatch in save_pretrained is unrelated to the generic bug report." }, { "left": "issue:43976", - "right": "issue:46218", + "right": "issue:44393", "accept": false, - "reason": "Python 3.9 compatibility and a Qwen3_5 import failure are unrelated." + "reason": "Python version compatibility and Qwen3-VL output hallucination are unrelated." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler inverse_sqrt kwargs issue and type-checking regression are not the same code path." + "reason": "Scheduler kwargs handling is unrelated to PretrainedConfig type checking." }, { "left": "issue:44492", "right": "issue:46328", "accept": false, - "reason": "Cache strategy typo and SequenceFeatureExtractor.pad performance issue are different problems." + "reason": "Cache-strategy typo and SequenceFeatureExtractor.pad inefficiency are different code paths." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-star performance regression and HybridCache import failure are distinct import-system issues, not duplicates." + "reason": "Import * performance regression is unrelated to missing HybridCache import/export." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor cached_file kwargs propagation and inverse_sqrt scheduler kwargs are unrelated." + "reason": "AutoProcessor cached_file kwargs forwarding is unrelated to inverse_sqrt scheduler kwargs." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import performance regression does not match a small testing_utils bug." + "reason": "Import performance regression is unrelated to a small testing_utils bug." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "Python compatibility regression and model import failure are different issues." + "reason": "Python-version incompatibility and module import failure are different issues." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init validation and _get_tied_weight_keys AttributeError are unrelated." + "reason": "post_init policy for patchtsmixer and _tied_weights_keys list handling are unrelated." } ] }, @@ -19190,16 +19152,16 @@ "issue:44485|issue:47176", "issue:43576|issue:44273", "issue:44360|issue:45468", - "issue:44315|issue:44829", "issue:43519|issue:46655", - "issue:44273|issue:44908", + "issue:44315|issue:44829", "issue:45230|issue:45310", + "issue:44273|issue:44908", "issue:44485|issue:47177", "issue:43519|issue:44485", "issue:45468|issue:46894", "issue:44655|issue:44861", - "issue:43576|issue:46218", "issue:43976|issue:47177", + "issue:43576|issue:46218", "issue:43519|issue:44908", "issue:41669|issue:43704", "issue:31795|issue:34689", @@ -19233,316 +19195,316 @@ "issue:44230|issue:44623", "issue:28348|issue:43352" ], - "prepared_review_unit_hash": "7241982cca639ba4f527fba4c76e44e4d5bb76aa3c650c61ad2899899909f9e0", + "prepared_review_unit_hash": "7a6bfce8ba2b99eae15f6c41210a40499eec99e940958a31467aa4dc5277ff85", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 47, - "serialized_chars": 40944, - "estimated_input_tokens": 10236, - "estimated_eval_tokens": 20728 + "serialized_chars": 40947, + "estimated_input_tokens": 10237, + "estimated_eval_tokens": 20730 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is very heterogeneous: most items are isolated model-loading, tokenizer, processor, scheduler, or security issues, with only a loose theme around v5 regressions. No soft-edge pair looks like the same underlying bug/change.", - "confidence": 0.78, - "canonical_issue_reason": "Issue 45850 is the broadest umbrella for the repeated \u201cv5 breaks models\u201d reports, so it is the best duplicate anchor even though it is still very generic.", + "summary": "The cluster is highly heterogeneous: most items are standalone Transformers bugs, regressions, or feature requests with only loose topical overlap (loading, tokenizer, model behavior, cache, security, etc.). No soft-edge pair looks like a true duplicate or a single fixable code-path conflict.", + "confidence": 0.86, + "canonical_issue_reason": "issue:33453 is the most broadly representative concrete bug report in this set (a model/tokenizer loading regression), so it is the least-bad anchor for a heterogeneous cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45850 is the most representative issue for the cluster\u2019s main regression theme; the rest are mostly narrower, unrelated, or model-specific bugs.", + "best_issue_reason": "issue:33453 is the strongest self-contained issue to use as a cluster anchor, but only as a representative loading-regression report, not as a true umbrella duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Tokenizer documentation confusion vs tokenizer loading regression; different symptom and fix." + "reason": "Both touch loading/docs around model inputs, but one is documentation confusion and the other is a tokenizer-loading regression; not the same bug." }, { "left": "issue:43976", "right": "issue:47176", "accept": false, - "reason": "Python-version compatibility bug vs shell-operator allowlist bypass; unrelated." + "reason": "Python-version support breakage and shell-operator allowlist bypass are unrelated issues." }, { "left": "issue:44485", "right": "issue:47176", "accept": false, - "reason": "GLM-5 RoPE implementation issue vs command-substitution security bypass; unrelated." + "reason": "GLM-5 RoPE implementation and command-substitution allowlist bypass are different code paths and different classes of bugs." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "Broken env CLI vs lazy-loading defect; different code paths." + "reason": "Broken `transformers env` CLI and lazy loading failure are unrelated behaviors." }, { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "DSA indexer activation bug vs Gemma audio positional-encoding issue; unrelated." - }, - { - "left": "issue:44315", - "right": "issue:44829", - "accept": false, - "reason": "model_init/Liger integration bug vs flash-attention training degradation; not the same bug." + "reason": "A DSA indexer ReLU discussion is not the same issue as Gemma-4 audio positional encoding." }, { "left": "issue:43519", "right": "issue:46655", "accept": false, - "reason": "Qwen3VL timestamp bug vs SmolVLM video padding bug; both processor-related but different defects." + "reason": "Qwen3VL timestamp math and SmolVLM video padding are both vision-related, but they are distinct processor bugs." }, { - "left": "issue:44273", - "right": "issue:44908", + "left": "issue:44315", + "right": "issue:44829", "accept": false, - "reason": "Lazy loading bug vs inverse_sqrt scheduler kwargs bug; unrelated." + "reason": "Liger Kernel not applied in `model_init` and flash_attention_3 degenerate training are separate training-path failures." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "Generic bug report vs Qwen3.5 from_pretrained error; too vague to match." + "reason": "Generic bug report versus a specific Qwen3.5 MoE `from_pretrained` regression; not a duplicate." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading malfunction and inverse_sqrt scheduler kwargs handling are unrelated." }, { "left": "issue:44485", "right": "issue:47177", "accept": false, - "reason": "GLM-5 RoPE issue vs tar/zip-slip RCE; unrelated." + "reason": "RoPE implementation bug and tar/zip slip RCE are completely different issues." }, { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Qwen3VL processor timestamp bug vs GLM-5 RoPE bug; unrelated." + "reason": "Qwen3VL timestamp handling and GLM-5 RoPE implementation are distinct model-specific bugs." }, { "left": "issue:45468", "right": "issue:46894", "accept": false, - "reason": "Gemma4 audio encoding bug vs CohereASR training-loss shift bug; different models and code paths." + "reason": "Gemma-4 audio positional encoding and CohereASR training-loss double-shift are different model behaviors, not the same underlying bug." }, { "left": "issue:44655", "right": "issue:44861", "accept": false, - "reason": "Pipeline save failure vs tied-weight-key AttributeError; different serialization bugs." + "reason": "Saving pipeline objects and `_get_tied_weight_keys` attribute errors are unrelated serialization/util bugs." }, { - "left": "issue:43576", - "right": "issue:46218", + "left": "issue:43976", + "right": "issue:47177", "accept": false, - "reason": "Broken env command vs Qwen3_5 import failure; unrelated." + "reason": "Python compatibility breakage and archive path traversal/RCE are unrelated." }, { - "left": "issue:43976", - "right": "issue:47177", + "left": "issue:43576", + "right": "issue:46218", "accept": false, - "reason": "Python-version support bug vs zip-slip/tar-slip RCE; unrelated." + "reason": "Broken env CLI and import failure for Qwen3_5ForCausalLM are different surfaces and different fixes." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor timestamp bug vs scheduler-kwargs bug; unrelated." + "reason": "Processor timestamp calculation and scheduler keyword handling are unrelated." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-star refactor/perf issue vs VRAM leak in Qwen3ForCausalLM; unrelated." + "reason": "Import-star performance issue and VRAM leak in Qwen3ForCausalLM are not the same bug." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Tokenizer docs issue vs Llama 3.2 model-loading regression; similar area but not the same bug." + "reason": "Documentation confusion about forward inputs and Llama 3.2 model-loading regression are different problems." }, { "left": "issue:45850", "right": "issue:47177", "accept": false, - "reason": "Broad v5 breakage complaint vs security file-write/RCE issue; unrelated." + "reason": "Broad version-breakage complaint and a command/file-write security issue are unrelated." }, { "left": "issue:45850", "right": "issue:47176", "accept": false, - "reason": "Broad v5 breakage complaint vs shell allowlist bypass; unrelated." + "reason": "General regression complaint and shell-operator allowlist bypass are not the same change or bug." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "GLM-5 RoPE bug vs Gemma4 audio positional-encoding bug; unrelated." + "reason": "Different model families and different math/encoding subsystems; not duplicates." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak vs RoPE implementation bug; unrelated." + "reason": "VRAM leak in Qwen3ForCausalLM and GLM-5 RoPE implementation are unrelated." }, { "left": "issue:46218", "right": "issue:47177", "accept": false, - "reason": "Import error vs tar/zip-slip RCE; unrelated." + "reason": "Import failure for Qwen3_5ForCausalLM and archive traversal/RCE are unrelated." }, { "left": "issue:46018", "right": "issue:46172", "accept": false, - "reason": "Shared-expert gating bug vs tensor transpose issue; unrelated." + "reason": "Shared-expert gating in DeepSeek-V4 and a `comb.to(...).transpose(...)` bug are distinct DeepSeek issues." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "local_files_only network leak vs processor timestamp bug; unrelated." + "reason": "Local-files-only network leakage and Qwen3VL timestamp calculation are unrelated." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 API/settings complaint vs VRAM leak; unrelated." + "reason": "TF32 API settings and Qwen3ForCausalLM VRAM leak are different subsystems." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "Broad v5 breakage complaint vs Qwen3_5 import error; unrelated." + "reason": "Broad version regression complaint and module import failure are not the same underlying issue." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Qwen3VL timestamp bug vs processor.save_pretrained missing files; different processor defects." + "reason": "Processor timestamp math and missing save_pretrained files are different processor/save issues." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "GlmMoeDsaIndexer split bug vs Gemma4 audio encoding issue; unrelated." + "reason": "torch.split return values in a DSA indexer and Gemma-4 audio positional encoding are unrelated." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "MLA rms_norm_eps precision bug vs Gemma4 audio encoding issue; unrelated." + "reason": "MLA layernorm epsilon precision bug and Gemma-4 audio positional encoding are different numerical issues." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Audio-model batching support issue vs DeepSeek weight-conversion question; unrelated." + "reason": "Continuous batching not supporting audio models and DeepSeek-V4 weight-conversion question are not the same bug." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Processor timestamp bug vs VRAM leak; unrelated." + "reason": "Qwen3VL timestamp calculations and VRAM leaks in Qwen3ForCausalLM are separate model-specific failures." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Nemotron FA2 support gap vs Qwen3_5 import error; unrelated." + "reason": "NemotronHForCausalLM Flash Attention 2 support and Qwen3_5ForCausalLM import failure are different issues." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "HybridCache import error vs Qwen3_5 import error; same broad symptom, but different missing symbols and fixes." + "reason": "HybridCache import failure and Qwen3_5ForCausalLM import failure are both import-related but concern different symbols and fixes." }, { "left": "issue:44861", "right": "issue:47110", "accept": false, - "reason": "Tied-weight-key AttributeError vs special_tokens list/keys error; unrelated." + "reason": "Tied-weights helper crash and special_tokens list/dict handling are different attribute/type bugs." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Both are tokenizer-v5 complaints, but they describe different failures and not clearly the same underlying bug." + "reason": "Tokenizer behavior change across v4/v5 and wrong tokenizer decoder type are related to tokenizers, but the reported failures and fixes are not clearly the same." }, { "left": "issue:44861", "right": "issue:47059", "accept": false, - "reason": "Tied-weight-key AttributeError vs huggingface_hub CLI completion error; unrelated." + "reason": "Tied-weights helper crash and CLI breakage with huggingface_hub are unrelated." }, { "left": "issue:28348", "right": "issue:45250", "accept": false, - "reason": "GPT2-specific flash-attention support request vs generic flash-attention topic; not the same concrete bug." + "reason": "GPT2 flash-attention support request and a generic Flash Attention 2.0 issue are too broad to treat as the same concrete change." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Scheduler kwargs bug vs unrelated placeholder issue." + "reason": "Scheduler kwargs bug and an unrelated placeholder issue title have no meaningful overlap." }, { "left": "issue:44998", "right": "issue:47177", "accept": false, - "reason": "Placeholder issue vs tar/zip-slip RCE; unrelated." + "reason": "Placeholder/unemployment issue and archive traversal/RCE are unrelated." }, { "left": "issue:44998", "right": "issue:47176", "accept": false, - "reason": "Placeholder issue vs shell allowlist bypass; unrelated." + "reason": "Placeholder/unemployment issue and shell-operator bypass are unrelated." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "Cache-strategy typo vs placeholder issue; unrelated." + "reason": "Cache-strategy typo and placeholder/unemployment issue are unrelated." }, { "left": "issue:44998", "right": "issue:45850", "accept": false, - "reason": "Placeholder issue vs broad v5 breakage complaint; unrelated." + "reason": "Placeholder/unemployment issue and broad version-breakage complaint are unrelated." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weight-key AttributeError vs assistant tool-call crash; unrelated." + "reason": "Tied-weight key crash and chat-template tool-call crash are different code paths." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Slow from_pretrained checkpoint loading vs missing layer_type_validation import; unrelated." + "reason": "Slow checkpoint loading and missing `layer_type_validation` import are unrelated import/loading problems." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 support request vs processor.save_pretrained missing files; unrelated." + "reason": "FP8 support request for qwen3-vl/qwen3.5 moe and missing processor save files are different feature/bug reports." }, { "left": "issue:28348", "right": "issue:43352", "accept": false, - "reason": "GPT2 FA2 support request vs Nemotron FA2 support gap; same broad topic, but different model-specific support gaps and not mergeable." + "reason": "Both mention Flash Attention 2.0, but they target different models and different support gaps; not a single fixable duplicate." } ] }, @@ -19604,14 +19566,14 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "b23d7d008025e2e550ee36281558ed861f4b4bd76e46c3d11a39e7bff6f3e217", + "prepared_review_unit_hash": "a73980f8c6fff4b30211f1b420f24abeedda712d84aeec25d235c7c9825a26f6", "component_budget": { "node_count": 517, "item_count": 517, - "soft_pair_count": 1906, - "serialized_chars": 988098, - "estimated_input_tokens": 247025, - "estimated_eval_tokens": 494306 + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 }, "budget": { "node_count": 17, @@ -19626,120 +19588,120 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated standalone issues across different models and subsystems. None of the soft pairs look like the same underlying bug; they only share broad themes like loading, model internals, or inference.", - "confidence": 0.91, - "canonical_issue_reason": "issue:29870 is the clearest concrete runtime bug report: it names a specific exception, a specific code path (`generate()`), and a reproducible failure mode. The others are either documentation, configuration, performance, or different model-specific problems.", + "summary": "This cluster is heterogeneous: it mixes unrelated bug reports, docs issues, config/type-hint nits, and model-specific feature requests. The soft pairs look like loose topical similarity rather than true duplicates, so none should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 43240 is the strongest standalone bug report in the set: it is concrete, code-path oriented, and has the most external discussion/inbound references. It is still not a true umbrella for the rest of the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:29870 is the strongest representative issue in this set because it is specific, actionable, and clearly describes a real execution bug rather than a vague enhancement or documentation concern.", + "best_issue_reason": "43240 is the best representative issue by activity/impact signals, but the overall set is too unrelated for a single duplicate canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:29870", "right": "issue:31515", "accept": false, - "reason": "Bart `generate()` UnboundLocalError and slow `from_pretrained` checkpoint loading are unrelated failure modes in different code paths." + "reason": "One is a BART generation UnboundLocalError; the other is a slow from_pretrained load report. Different failures and code paths." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Checkpoint load speed and tensor-parallel training correctness are different problems with no shared code-path evidence." + "reason": "Checkpoint loading slowness and tensor-parallel training bugs are unrelated symptoms, not the same underlying defect." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow reporting on failed training jobs is unrelated to checkpoint loading performance." + "reason": "MLflow reporting after failed jobs is unrelated to model loading performance." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "A loading-performance complaint is not the same as confusing forward-API documentation." + "reason": "Performance complaint about loading checkpoints vs a documentation confusion issue; not the same bug." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "Both involve Qwen3.5/MoE, but one is fp8 inference support and the other is a `from_pretrained` failure; not clearly the same bug." + "reason": "Both mention Qwen3.5/MoE and fp8/from_pretrained, but one is an fp8 inference support request and the other is a version-specific load error; not clearly the same fix." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Continuous batching for audio models and Qwen3.5 fp8 inference support are distinct feature gaps." + "reason": "Continuous batching audio support and fp8 Qwen VL/MoE inference support are different feature gaps." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba/Mamba-2 init-weights behavior and RoFormer attention-interface usage are unrelated implementation issues." + "reason": "Mamba init-weights behavior and RoFormer attention-interface usage are separate model implementation issues." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former `ignore_value` handling and DeepSeek-V4 shared-expert gating are different model-specific bugs." + "reason": "Mask2Former ignore_value handling and DeepSeek-V4 shared-expert gating concern different models and different logic." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "A config type-hint mistake and tokenizer regex warnings after a version upgrade are not the same underlying issue." + "reason": "A config type-hint typo is not the same as a tokenizer regex warning regression." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Unused `ignore_value` logic and missing kwargs passed to loss calculation are different bugs." + "reason": "Mask2Former ignore_value usage and loss-kwargs propagation are different code paths and fixes." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss-function kwargs handling and RoFormer attention interface usage do not overlap in code path or symptom." + "reason": "Loss-function kwargs forwarding and RoFormer attention plumbing are unrelated changes." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former loss-parameter handling is unrelated to tokenizer warning behavior." + "reason": "Model loss/ignore_value behavior and tokenizer warning behavior do not share the same underlying bug." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "These affect different models and different subsystems; no evidence of a shared defect." + "reason": "Different model families and different implementation defects; not mergeable as one issue." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Unused loss config and a typo in a quantization string are unrelated." + "reason": "Mask2Former loss handling and a typo in quantization text are unrelated." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Incorrect type hints in ModernBERTConfig are not the same as a typo in conversion code." + "reason": "Incorrect type hint versus typo in an attribute path; these are separate nits." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Missing kwargs in loss calculation and a config type-hint issue are different categories of bug." + "reason": "Loss kwargs forwarding and a config annotation typo are not the same defect." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "A design/question-about-norm implementation and a type-hint correction are not duplicate reports." + "reason": "A design question about RMS norm and a type-hint correctness issue are unrelated." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former loss handling and ModernBERT config typing are unrelated." + "reason": "Mask2Former ignore_value behavior and ModernBERT config typing are separate issues with no shared fix." } ] }, @@ -19772,12 +19734,12 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "df3f2797d61f23ac7e6e80f1f24f86dfdcd704416f2220ee0171dd25de4675f3", + "prepared_review_unit_hash": "03819896d0fefa47e0321e9b2942a9f4fa741d5de4c822095d5490bfedf30303", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2874, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -19785,7 +19747,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2874, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -19794,30 +19756,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues are unrelated feature requests: token truncation direction, ReduceLROnPlateau-like scheduling behavior, and Wav2Vec2 word boundary time offsets. None appear to describe the same underlying bug or change, so no duplicates should be merged.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 10519 is the clearest, most specific request with a well-defined API change; the other issues cover different models and unrelated functionality.", + "summary": "These three issues are about unrelated feature requests: token truncation behavior, ReduceLROnPlateau-like training behavior, and Wav2Vec2 word time offsets. They do not describe the same underlying bug or change, so none should be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "10519 is the best representative issue because it is concrete, narrowly scoped, and easier to resolve than the more open-ended scheduler and Wav2Vec2 questions.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different domains and requests: truncation behavior vs. extracting word time offsets in Wav2Vec2." + "reason": "Different areas and requests: tokenizer truncation direction vs Wav2Vec2 word boundary timestamps. No shared underlying bug/change." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "One asks for tokenizer truncation options; the other asks for ReduceLROnPlateau-like learning rate scheduling. No shared underlying change." + "reason": "Tokenizer truncation option request is unrelated to a ReduceLROnPlateau-like scheduling feature request." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "Unrelated functionality: learning-rate scheduling vs. Wav2Vec2 alignment/time-offset extraction." + "reason": "Training-scheduler functionality and Wav2Vec2 time-offset extraction are unrelated feature requests, not the same fix." } ] }, @@ -19849,12 +19811,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "76b19b55d55dcd5bf43ef24913ed28df553b440c0bfccacacdc3879cc9f57e42", + "prepared_review_unit_hash": "90715683bc6f6889882af4444cbcd4eb0263f75eb5cf7cfbbb79701201107945", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2478, + "serialized_chars": 2479, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -19862,7 +19824,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2478, + "serialized_chars": 2479, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -19871,24 +19833,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The three issues only weakly resemble each other at a very broad NLP/model-loading level, but they describe different problems: tokenization mismatch, a private-model authentication/error-message issue, and a specific failure loading bert-base-uncased. They do not look like duplicates of one underlying bug.", - "confidence": 0.93, - "canonical_issue_reason": "Issue #16618 is the broadest and most central-sounding of the three, with the most discussion and a direct model-loading failure title, so it is the best representative if one must be chosen.", + "summary": "These issues are not duplicates of each other. Two are about model-loading failures, but they point to different models and root causes; the private-model auth message issue is clearly separate. Issue 16618 is the broadest user-facing failure and the best cluster representative.", + "confidence": 0.86, + "canonical_issue_reason": "#16618 is the most general and central user-facing failure in the set: a failed model load for a common model name, with the broadest applicability and strongest discussion activity.", "canonical_pr_reason": null, - "best_issue_reason": "#16618 is the best single issue to anchor the cluster because it is the most general model-loading report and has the highest discussion activity.", + "best_issue_reason": "#16618 best represents the cluster because it is the broadest loading-failure report, while the others are narrower and point to different causes.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tapas tokenization differing from TensorFlow code is a tokenizer/processing discrepancy, while the other issue is about loading bert-base-uncased. They concern different failure modes and code paths." + "reason": "Different underlying problems: Tapas tokenization mismatch versus failure to load bert-base-uncased. The titles do not indicate the same code path or fix." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "A bad error message when downloading a private model without login is an auth/error-handling problem, not the same as an inability to load bert-base-uncased." + "reason": "Different root cause and symptom: a bad auth/error message when downloading a private model versus an inability to load a public model. Not the same bug." } ] }, @@ -19918,40 +19880,40 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "f5714d5029d015d5eb1e031def1de875b6f396f09ffd4159754982b5d66ac295", + "prepared_review_unit_hash": "73f15ea417e2167142b5d8aa8796047069ff9d45d36023b6735428e0fef96bee", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1587, - "estimated_input_tokens": 397, - "estimated_eval_tokens": 1050 + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1587, - "estimated_input_tokens": 397, - "estimated_eval_tokens": 1050 + "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 two issues are not duplicates: one requests Flash Attention 2 support across more architectures, while the other requests GGUF support across more architectures. They share only a broad phrasing pattern, not the same feature or code change.", - "confidence": 0.98, + "summary": "These are two separate community feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They are only superficially similar and should not be treated as duplicates.", + "confidence": 0.99, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a good canonical representative because the requests concern different subsystems and different feature work; the similarity is only superficial.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Reject: Flash Attention 2 architecture support and GGUF architecture support are distinct feature requests with different implementation paths and targets." + "reason": "Different underlying requests and code paths: Flash Attention 2 support vs GGUF support. They share a broad theme of expanding architecture support, but not the same bug/change." } ] }, @@ -20016,12 +19978,12 @@ "issue:46222|issue:46931", "issue:46226|issue:46931" ], - "prepared_review_unit_hash": "0704df30e24d00def90f104ca3fa31728aa8012c5f8fe81eac914d22783aedff", + "prepared_review_unit_hash": "4dc9f03ea018c810bb421baed53321d50ca8c39f3c27450d95cf848808c7f531", "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 }, @@ -20029,7 +19991,7 @@ "node_count": 9, "item_count": 9, "soft_pair_count": 29, - "serialized_chars": 15983, + "serialized_chars": 15982, "estimated_input_tokens": 3996, "estimated_eval_tokens": 8248 }, @@ -20038,195 +20000,195 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Mostly separate i18n translation requests rather than one duplicate bug; only the near-identical template issues and the exact same-title Thai pair look mergeable.", - "confidence": 0.66, - "canonical_issue_reason": "issue:43834 is the best representative of the largest near-identical generic translation-request subgroup: it is earliest, has the broadest overlap with close matches, and looks like the central duplicate anchor.", + "summary": "The cluster splits into two clear groups: distinct language-specific translation requests, and a duplicate-like placeholder-template subgroup with identical titles and near-identical body sizes. The placeholder subgroup looks mergeable as duplicates, but I would not treat the language-specific requests as duplicates of those template reports. The two Thai-title issues form a separate exact-duplicate pair; the different Thai-titled issue remains separate.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43834 is the best representative of the repeated placeholder-title subgroup: it is the earliest item in the largest same-title cluster and most cleanly captures the malformed template pattern.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43834 is the strongest global issue representative because it sits in the densest cluster of nearly identical titles/bodies and is the clearest canonical anchor among the issue duplicates.", + "best_issue_reason": "issue:43834 is the strongest overall issue representative because it anchors the largest evident duplicate-like cluster, while the language-specific translation requests are still distinct from each other.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different target-language translation request; similar template text is not enough to make these the same underlying issue." + "reason": "Turkish translation request vs placeholder-title report; the titles indicate different concrete issues." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different target-language translation request; related i18n topic, but not the same change." + "reason": "Turkish translation request vs placeholder-title report; similar boilerplate is not enough to merge them." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Different language request; same translation-doc theme, but not the same underlying issue." + "reason": "Turkish language-specific request and placeholder-template report are not the same underlying issue." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Different language request; should not be merged as a duplicate." + "reason": "Different requested languages (Turkish vs Romanian), so these are separate translation issues." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish vs Spanish translation requests are distinct work items, not duplicates." + "reason": "Different requested languages (Turkish vs Spanish), so these are distinct issues." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Different language request; similarity is only at the template level." + "reason": "Spanish translation request vs placeholder-template report; no evidence they target the same concrete task." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Different language request; same broad subsystem, but not the same change." + "reason": "Spanish translation request vs placeholder-template report; only the template structure overlaps." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Different language request; not the same underlying issue." + "reason": "Spanish and Romanian translation requests are separate issues." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Different language request; similar template body does not make it a duplicate." + "reason": "Spanish translation request vs placeholder-template report; not the same underlying issue." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian translation request is distinct from the generic template issues." + "reason": "Romanian translation request vs placeholder-template report; similar formatting does not make them duplicates." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Different language request; not the same underlying change." + "reason": "Romanian translation request vs placeholder-template report; not enough evidence of the same issue." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Different language request; the overlap is only in the translation-doc template." + "reason": "Romanian translation request and placeholder-template report are different concrete issues." }, { "left": "issue:38435", "right": "issue:46222", "accept": false, - "reason": "The titles do not indicate the same concrete request; body similarity alone is too weak here." + "reason": "Romanian doc-translation request and the Thai-titled report are unrelated concrete issues." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "The titles do not indicate the same concrete request; too much ambiguity to merge." + "reason": "Romanian doc-translation request and the Thai-titled report are unrelated concrete issues." }, { "left": "issue:43834", "right": "issue:44370", "accept": true, - "reason": "Same generic title/template and near-identical body; looks like a duplicate translation-request issue." + "reason": "Identical placeholder-style title and nearly identical body size; this looks like a repeated malformed-template submission." }, { "left": "issue:43834", "right": "issue:44365", "accept": true, - "reason": "Same generic title/template and near-identical body; very likely the same issue reported twice." + "reason": "Identical placeholder-style title and nearly identical body size; strong duplicate signal." }, { "left": "issue:43834", "right": "issue:46222", "accept": false, - "reason": "Titles are not the same and the visible topic differs; template similarity alone is insufficient." + "reason": "Different title and different apparent subject; soft similarity alone is not enough to merge them." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Titles are not the same and the visible topic differs; do not merge on body similarity alone." + "reason": "Different title and different apparent subject; these do not clearly describe the same issue." }, { "left": "issue:43834", "right": "issue:46931", "accept": false, - "reason": "Different title/topic; only the template text looks similar, which is not enough." + "reason": "Different Thai title/topic; there is no concrete evidence these are the same report." }, { "left": "issue:44365", "right": "issue:44370", "accept": true, - "reason": "Same generic title/template with highly similar bodies; this looks like a duplicate." + "reason": "Same placeholder title and almost identical body size; they look like duplicate malformed-template reports." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Different visible title/topic; not enough evidence that this is the same request." + "reason": "Placeholder-template report vs Thai-titled report; the titles do not indicate the same underlying issue." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Different visible title/topic; body similarity alone is too ambiguous." + "reason": "Placeholder-template report vs Thai-titled report; not enough evidence of duplication." }, { "left": "issue:44365", "right": "issue:46931", "accept": false, - "reason": "Different topic; reject despite generic-body similarity." + "reason": "Different Thai title/topic; similarity is only superficial." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Different visible title/topic; should not be merged as the same issue." + "reason": "Placeholder-template report vs Thai-titled report; no clear common concrete issue." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Different visible title/topic; too ambiguous to accept." + "reason": "Placeholder-template report vs Thai-titled report; not enough evidence to merge." }, { "left": "issue:44370", "right": "issue:46931", "accept": false, - "reason": "Different topic; not the same underlying issue." + "reason": "Different Thai title/topic; the similarity score alone is not sufficient." }, { "left": "issue:46222", "right": "issue:46226", "accept": true, - "reason": "Exact same title and same-sized body; this is the clearest duplicate pair in the set." + "reason": "Same exact Thai title and identical body length; this is a strong duplicate pair." }, { "left": "issue:46222", "right": "issue:46931", "accept": false, - "reason": "Different title/topic; no evidence of the same underlying request." + "reason": "Different Thai title and no explicit linkage; high similarity alone does not prove duplication." }, { "left": "issue:46226", "right": "issue:46931", "accept": false, - "reason": "Different title/topic; the similarity appears incidental." + "reason": "Different Thai title and no explicit linkage; not clearly the same concrete report." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary stays grounded in the packet: the Turkish/Spanish/Romanian items are clearly distinct language-specific requests, the three generic i18n template issues are plausibly duplicate-like based on identical titles/high similarity, and the Thai pair has identical titles and matching body length. The soft-edge verdicts are conservative overall, with no obvious overstatement beyond what the packet supports." + "accept": false, + "feedback": "The high-level separation between the language-specific translation requests and the placeholder-template reports is plausible, but the summary overstates the evidence for the placeholder group by calling them exact duplicates. The packet only gives titles, body lengths, and soft-similarity scores; it does not show the bodies themselves. Also, the pairwise verdicts are internally inconsistent: 43834\u219444365 and 44365\u219444370 are accepted, but 43834\u219444370 is rejected despite even stronger similarity. That makes the duplicate-subgroup claim too strong for the available evidence." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -20316,12 +20278,12 @@ "issue:42503|issue:46653", "issue:45027|issue:46653" ], - "prepared_review_unit_hash": "146dfe940068631094e105cfcac1bf7db19e29f24892803b0e0a17eba390a741", + "prepared_review_unit_hash": "f52eab939b7c2da6741174d092e5192e5147e3615e0f0f313b7bcaa814cd41d4", "component_budget": { "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28617, + "serialized_chars": 28620, "estimated_input_tokens": 7155, "estimated_eval_tokens": 14566 }, @@ -20329,7 +20291,7 @@ "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28617, + "serialized_chars": 28620, "estimated_input_tokens": 7155, "estimated_eval_tokens": 14566 }, @@ -20338,96 +20300,96 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These issues are mostly unrelated feature requests spanning different model families and tasks (keypoint matching, BERT variants, detection, audio ASR/TTS, OCR, etc.). The apparent similarity is largely templated wording like \"Add/Support model,\" not a shared underlying bug or change, so this is not a duplicate cluster.", - "confidence": 0.95, - "canonical_issue_reason": "No single canonical issue stands out: the items describe different models, modalities, and code paths rather than one underlying request.", + "summary": "This cluster is mostly a grab bag of unrelated model-addition requests across vision, audio, and NLP. There are naming/theme overlaps, but no pair looks like the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.87, + "canonical_issue_reason": "No single issue is a true duplicate target; if one issue must anchor the cluster, issue 36879 is the broadest open model-add request and sits near the center of the similarity graph, but it is still only a representative.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong global representative for duplicate triage because the cluster is too heterogeneous; any pick would overstate duplication.", + "best_issue_reason": "Issue 36879 is the most generic and central open request in this set, but the cluster spans multiple unrelated model families, so it is not a definitive canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36718", "right": "issue:47004", "accept": false, - "reason": "Both are keypoint-related, but they request different models/features (RoMa matcher vs LoMa detector+matcher), so not the same change." + "reason": "RoMa vs LoMa are different requested keypoint components; related naming, not the same change." }, { "left": "issue:36718", "right": "issue:41115", "accept": false, - "reason": "Different domains: keypoint matching versus adding a MiniCPM3 architecture." + "reason": "Different model families and tasks: keypoint matching vs MiniCPM3 architecture support." }, { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config and DEIMv2 are unrelated model additions." + "reason": "Both are model requests, but for different models with no shared concrete implementation target." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert configuration work is unrelated to Qwen3-TTS support." + "reason": "EuroBert config support and Qwen3-TTS support are unrelated changes." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "These are separate model requests for different architectures (EuroBert vs RF-DETR)." + "reason": "Different model requests: EuroBert config vs RF-DETR model addition." }, { "left": "issue:36817", "right": "issue:42954", "accept": false, - "reason": "EuroBert and XiaoMi MiMo are distinct model families with no shared code-path issue." + "reason": "EuroBert config and MiMo are distinct model additions." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are different vision models, not one duplicate request." + "reason": "RF-DETR and DEIMv2 are different vision model additions." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "Object detection (RF-DETR) and audio model support (LFM2.5 Audio) are unrelated." + "reason": "RF-DETR and LFM2.5 Audio are different modalities and code paths." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR and a weight-sparse transformer request are different model changes." + "reason": "RF-DETR model support and weight-sparse transformer support are different changes." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR versus Qwen3-ASR are different tasks and code paths." + "reason": "RF-DETR and Qwen3-ASR are unrelated model support requests." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR and Qwen3-TTS are unrelated model support requests." + "reason": "RF-DETR vision model support is not the same change as Qwen3-TTS support." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are distinct model requests with different implementations." + "reason": "NeoBERT and Qwen3-ASR are different model families and tasks." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT versus a sparse-transformer request is not a duplicate." + "reason": "NeoBERT addition and sparse-transformer support are separate implementation efforts." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT and LFM2.5 Audio target different modalities and code paths." + "reason": "NeoBERT and LFM2.5 Audio target different architectures and modalities." }, { "left": "issue:36879", @@ -20439,7 +20401,7 @@ "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT models are different requests, not one underlying change." + "reason": "NeoBERT and ModernVBERT are different model requests, not duplicates." }, { "left": "issue:41115", @@ -20451,169 +20413,199 @@ "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 support is not the same as adding a sparse-transformer model." + "reason": "MiniCPM3 and weight-sparse transformer support do not describe the same concrete fix." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different model families." + "reason": "MiniCPM3 architecture and ModernVBERT are distinct model additions." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 architecture support and Voxtral TTS support are unrelated." + "reason": "MiniCPM3 and Voxtral TTS support are different model integration tasks." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are different model requests with no shared duplicate scope." + "reason": "DEIMv2 and Qwen3-TTS are unrelated model support requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT models and Qwen3-ASR support are unrelated changes." + "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 different model families and tasks." + "reason": "ModernVBERT and Qwen3-TTS are not the same change." }, { "left": "issue:42503", "right": "issue:42954", "accept": false, - "reason": "ModernVBERT models and XiaoMi MiMo are separate requests." + "reason": "ModernVBERT and MiMo are separate model additions." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "A BERT-like RoPE request and Voxtral TTS support do not share the same underlying change." + "reason": "BERT-like RoPE support and Voxtral TTS support are unrelated requests." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models and BERT-like models with RoPE are related only by broad theme, not by the same concrete request." + "reason": "ModernVBERT and BERT-like RoPE models are related only broadly, not as the same concrete change." }, { "left": "issue:42738", "right": "issue:42954", "accept": false, - "reason": "BERT-like RoPE support and XiaoMi MiMo are unrelated model additions." + "reason": "BERT-like RoPE support and MiMo are different model requests." }, { "left": "issue:42738", "right": "issue:46653", "accept": false, - "reason": "BERT-like model support and a personal-name issue are clearly unrelated." + "reason": "BERT-like RoPE support and Mirko Privitera are unrelated." }, { "left": "issue:42954", "right": "issue:44205", "accept": false, - "reason": "XiaoMi MiMo and SAM3-LiteText are different models and different code paths." + "reason": "MiMo and SAM3-LiteText are different model additions." }, { "left": "issue:42954", "right": "issue:43671", "accept": false, - "reason": "XiaoMi MiMo and Qwen3-TTS are separate model support requests." + "reason": "MiMo and Qwen3-TTS are unrelated model support requests." }, { "left": "issue:42954", "right": "issue:45177", "accept": false, - "reason": "XiaoMi MiMo and DeepSeek-OCR2 are unrelated model additions." + "reason": "MiMo and DeepSeek-OCR2 are different model requests." }, { "left": "issue:42954", "right": "issue:46653", "accept": false, - "reason": "XiaoMi MiMo and the Mirko Privitera issue are not the same underlying bug/change." + "reason": "MiMo and Mirko Privitera are unrelated; one is a model request, the other is not." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are same vendor but different modalities/tasks; they would not merge into one PR safely." + "reason": "Qwen3-TTS and Qwen3-ASR are same vendor family but different concrete capabilities and code paths." }, { "left": "issue:43671", - "right": "issue:45027", + "right": "issue:44121", "accept": false, - "reason": "Both are TTS-related, but they target different models and implementation paths." + "reason": "Qwen3-TTS and sparse-transformer support are not the same fix." }, { "left": "issue:43671", - "right": "issue:45177", + "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS support and DeepSeek-OCR2 are unrelated." + "reason": "Qwen3-TTS and Voxtral TTS are both audio-related, but they target different models and implementations." }, { "left": "issue:43671", - "right": "issue:44121", + "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and a sparse-transformer model request are different changes." + "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model requests." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are separate audio-model support requests, not one concrete fix." + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio model integrations." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR and EUPE are unrelated model requests." + "reason": "Qwen3-ASR and EUPE are unrelated model additions." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR support does not overlap with adding a weight-sparse transformer." + "reason": "Qwen3-ASR support and sparse transformer support are different concrete changes." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are different audio models and tasks." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 are unrelated." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are different model requests." + }, + { + "left": "issue:43909", + "right": "issue:46645", + "accept": false, + "reason": "LFM2.5 Audio support and Granite SWA+sink option are not the same code-path change." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Sparse transformer support and EUPE model addition are unrelated requests." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "ASR and TTS are different tasks, and these are different model families." + "reason": "Qwen3-ASR and Voxtral TTS are both speech-related, but they are distinct model integrations." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS are unrelated model additions." + "reason": "NeoBERT and Voxtral TTS are unrelated model requests." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR and DeepSeek-OCR2 are different model families and tasks." + "reason": "Qwen3-ASR and DeepSeek-OCR2 target different modalities and implementations." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are separate model requests, not the same underlying change." + "reason": "EUPE and TIPSv2 are both vision-language related, but they are different model additions." }, { "left": "issue:42503", "right": "issue:46653", "accept": false, - "reason": "ModernVBERT models and the Mirko Privitera issue are unrelated." + "reason": "ModernVBERT and Mirko Privitera are unrelated." }, { "left": "issue:45027", "right": "issue:46653", "accept": false, - "reason": "Voxtral TTS support and the Mirko Privitera issue are clearly unrelated." + "reason": "Voxtral TTS support and Mirko Privitera are unrelated; no shared code change." } ] }, @@ -20643,7 +20635,7 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "eef7196dd5804f53be4d77246eafb40ce5ac1a241a73744ee9983c2c05f180b5", + "prepared_review_unit_hash": "631a614f3578018bd0c27d8cda67f808a10e43abdd724f6abe2d5ffbacc34e81", "component_budget": { "node_count": 2, "item_count": 2, @@ -20665,18 +20657,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are unrelated: one is a docs request to add Pruna as an optimization framework, and the other is a Vision RoPE refactor. There is no duplicate relationship here.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue: these are different topics in different areas, with no shared targets or concrete overlap.", + "summary": "These two issues do not appear to describe the same underlying problem: one is a docs request about adding Pruna as an optimization framework, while the other is an open technical refactor for Vision RoPE. They should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue for the pair: the topics, scopes, and expected fixes are unrelated (documentation/integration request vs. model architecture refactor).", "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good global anchor for this cluster because they do not describe the same underlying bug or change.", + "best_issue_reason": "Issue #46443 is the better representative of the cluster only because it is a substantive code-related issue; #38740 is a documentation request and is not a duplicate partner for the refactor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Reject: a documentation/framework listing request is not the same underlying change as a Vision RoPE refactor, and there is no shared code path or concrete overlap." + "reason": "Different subjects and remediation paths: a docs request to add Pruna versus a Vision RoPE refactor. No shared code path or concrete fix target is evident." } ] }, @@ -20706,7 +20698,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "9908cbdf95831c8a8c28945feeec8bf4c9d5ecac1405da1dd0e85de6af436f41", + "prepared_review_unit_hash": "ee37ce541e43b85acad762757e3a4ae2eaf5b5275e723f30f19722852bf64805", "component_budget": { "node_count": 2, "item_count": 2, @@ -20728,18 +20720,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one is a vague request about block decoding, while the other is a model-specific feature request for DeepSeek v3.2 support. They may both relate to inference/backend work, but they describe different underlying changes.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue: the requests target different features and are not the same bug or change. If one must be chosen as the cluster representative, 43037 is clearer and more specific.", + "summary": "These two issues are related only loosely by NLP/LLM inference context, but they describe different work: one is about block decoding mechanics, the other is a model-specific DeepSeek v3.2 support request. I would not triage them as duplicates.", + "confidence": 0.89, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43037 is the better representative because it is a concrete, self-contained support request with a specific target (DeepSeek v3.2), whereas 41084 is broader and less explicit.", + "best_issue_reason": "#43037 is the better standalone issue because it is a concrete feature request with a clear target (DeepSeek v3.2 support), while #41084 is more ambiguous and implementation-oriented.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different requests: 'Set Block Decoding' and 'DeepSeek v3.2 support' do not describe the same underlying bug or change, and there is no shared target or code path evidence." + "reason": "Reject: the titles point to different underlying changes. \"Set Block Decoding\" sounds like an internal decoding-mode/algorithm setting, while \"DeepSeek v3.2 support\" is model compatibility work. There is no clear shared bug or code path, so they should not be merged as duplicates." } ] }, @@ -20769,12 +20761,12 @@ "soft_pairs": [ "issue:41306|issue:41826" ], - "prepared_review_unit_hash": "0a5ee9021e0420b3fbb5ca1d5d77ce018b1143613a39fd879e166a42ee83cbfe", + "prepared_review_unit_hash": "a5d29226d408e70c3d8fe504a842d4b0549dfc75a07047f44db7a63a958704d8", "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 }, @@ -20782,7 +20774,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -20791,8 +20783,8 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are both about memory-optimization ideas, but they target different components and different changes: one requests a `from_pretrained()` flag for TiledFusedLogitsLoss/FLCE, the other asks to integrate TiledMLP. They do not look like the same underlying bug or feature request.", - "confidence": 0.96, + "summary": "These two issues are about different feature requests in different parts of the model stack: one for a TiledFusedLogitsLoss / Liger FLCE flag in `from_pretrained()`, the other for integrating TiledMLP to reduce memory footprint. They do not look like the same underlying bug or change.", + "confidence": 0.94, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, @@ -20802,7 +20794,7 @@ "left": "issue:41306", "right": "issue:41826", "accept": false, - "reason": "Different requested features in different code paths (logits loss flag vs TiledMLP integration); shared memory-footprint theme is too broad to treat as the same issue." + "reason": "Different requested capabilities and code paths; only broad memory/performance theme overlaps, which is not enough to treat them as duplicates." } ] }, @@ -20832,7 +20824,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "79c3f02c646fc6ce3ccb57c2bb1563d2c9af4e8babab9df644d0ec8abf466508", + "prepared_review_unit_hash": "14257f0c1f57a5d4c0a50095a85263bbed254dba8e62c4b10bbf88441bd2b6e7", "component_budget": { "node_count": 2, "item_count": 2, @@ -20854,18 +20846,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues both mention custom trainer extensibility, but they are about different behaviors: one is about `remove_unused_column` handling, the other about FLOPS estimation for custom trainers. They are not the same underlying bug or change.", - "confidence": 0.91, - "canonical_issue_reason": "Issue #43598 is the more concrete and code-specific request, centered on `Trainer.remove_unused_column` customizability.", + "summary": "The two issues are related to custom trainers, but they describe different problems in different Trainer code paths: unused-column handling vs FLOPS estimation.", + "confidence": 0.87, + "canonical_issue_reason": "Issue #43598 is the more concrete and representative report, focused on a specific Trainer behavior (`remove_unused_column`) and its impact on customizability.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43598 is the better representative issue because it identifies a specific Trainer behavior and likely actionable code path.", + "best_issue_reason": "#43598 is the better cluster representative because it names the specific mechanism being questioned and reads as the more actionable root concern.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Related only at a high level (custom trainer support), but they target different functionality: column removal vs FLOPS estimation." + "reason": "They are both about custom trainers, but the underlying bugs differ: one concerns `remove_unused_column` behavior, the other concerns how FLOPS are estimated for custom trainer classes. Different fixes/code paths." } ] }, @@ -20895,7 +20887,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "dd564f7bf09c999273e3261896cb91b789287105a91c871ba84fc847a020602a", + "prepared_review_unit_hash": "50204f3d80fef2dc2638e429a08b9730b6a0c372e73f7bd06d87d3f461d92049", "component_budget": { "node_count": 2, "item_count": 2, @@ -20917,18 +20909,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are only loosely related by theme (Qwen auto-class support) but target different model families and different features: one is about registering Qwen3-Omni for AutoModel/AutoModelForConditionalGeneration, the other is about adding AutoModelForSequenceClassification support for Qwen3.5. They are not duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "If one must be chosen, issue 43684 is the better canonical because it is broader and more central to auto-class registration support, while 44405 is a separate task-specific request.", + "summary": "These are related only at a broad transformer-automation level, but they request different model families and different AutoModel heads, so they should not be clustered as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #43684 is the broader, more foundational registration request: adding Qwen3-Omni to AutoModel and AutoModelForConditionalGeneration.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the strongest representative of the cluster because it concerns a generic registration gap in the auto-model APIs; however, the pair is not truly duplicate-worthy.", + "best_issue_reason": "Issue #43684 is the better representative because it covers a core model registration gap rather than a narrower head-specific feature request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying changes: Qwen3-Omni auto-model registration vs Qwen3.5 sequence-classification support. They share a broad ecosystem but not the same concrete bug or mergeable fix." + "reason": "Different underlying requests: Qwen3-Omni model registration for AutoModel/AutoModelForConditionalGeneration vs Qwen3.5 AutoModelForSequenceClassification support. Same ecosystem, but not the same bug or change." } ] }, @@ -20965,79 +20957,79 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "fc3ac903d3f4bd3c381932e6c4edd6cc23f6650fe2fa6e3f8dc2567f09ee5e91", + "prepared_review_unit_hash": "e31cbccf2e53bff336b25064158fe7cde62f2f4f0d09dba2b119668581ae51f7", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4262, - "estimated_input_tokens": 1066, - "estimated_eval_tokens": 2388 + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4262, - "estimated_input_tokens": 1066, - "estimated_eval_tokens": 2388 + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Issues 43845, 43846, and 43859 share the same title ('huggingface ') and are plausible duplicates; 43835 ('Fett') looks separate and should not be merged into that cluster.", - "confidence": 0.76, - "canonical_issue_reason": "Issue 43859 is the strongest canonical pick within the 'huggingface ' group because it shares the exact title with the other two and is the latest of the three, with some discussion activity.", + "summary": "Three issues share the same generic title \"huggingface \" and are likely duplicates of one another; \"Fett\" appears unrelated. I\u2019d keep issue 43845 as the canonical issue among the huggingface trio because it is the earliest and most representative entry.", + "confidence": 0.72, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface \" reports and is a clean representative for the duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43859 is the best issue candidate for the 'huggingface ' cluster since it has the same title as the others, is the most recent, and has a bit more activity than the other two.", + "best_issue_reason": "43845 is the best issue to anchor the cluster: earliest timestamp among the near-identical huggingface reports, with the other two likely dupes.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles ('Fett' vs 'huggingface ') and no other evidence tying them to the same report." + "reason": "Different titles and no supporting detail tying \"Fett\" to the huggingface reports." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no concrete evidence that they describe the same issue." + "reason": "Different titles and no evidence of the same underlying bug or request." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "'Fett' does not match the 'huggingface ' reports, so there is no basis to treat them as duplicates." + "reason": "Different titles; nothing indicates the same issue as the huggingface reports." }, { "left": "issue:43845", "right": "issue:43846", "accept": true, - "reason": "Same title and nearly the same creation time, which is consistent with duplicate reports." + "reason": "Same title and highly similar sparse content; these look like duplicate reports of the same issue." }, { "left": "issue:43845", "right": "issue:43859", "accept": true, - "reason": "Same title, with 43859 appearing as another report of the same issue despite being later." + "reason": "Same title and clearly overlapping generic report; likely the same underlying issue." }, { "left": "issue:43846", "right": "issue:43859", "accept": true, - "reason": "Same title and the same soft similarity signal as the other 'huggingface ' pairs, so they look duplicate-like." + "reason": "Same title and near-identical high-level report; plausible duplicates of the same issue." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The clustering is partly grounded, but several claims overstate the packet. It is safe to say the three 'huggingface ' issues share the same title and that issue:43835 ('Fett') is different. However, the canonical/best-issue reasons go beyond the evidence by calling the other reports 'essentially identical placeholders' and 'same placeholder-style content'\u2014the packet only shows title similarity, timestamps, and minimal activity, not body content or placeholder semantics. Also, 'closely related timing' is a bit strong for issue:43859 versus the earlier two. Please keep the duplicate rationale limited to the observed soft similarity and avoid asserting identical underlying content." + "accept": true, + "feedback": "Grounded and conservative: 43845 is the earliest of the repeated \"huggingface \" issues, 43835 has a different title and no supporting link, and the duplicate verdicts among 43845/43846/43859 are supported by the exact title match plus the soft-similarity evidence. No overstatement beyond the packet." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -21076,82 +21068,82 @@ "issue:44573|issue:47036", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "6797cdb025b087baf4bbbd7b7c600556578510fc1795207f4744aba9543566f3", + "prepared_review_unit_hash": "cac0291dbcb9a03e6ee019ea7f265a9c8c52d60c1d8c9afaccba26dd39215a2b", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7349, - "estimated_input_tokens": 1838, - "estimated_eval_tokens": 3932 + "serialized_chars": 7358, + "estimated_input_tokens": 1840, + "estimated_eval_tokens": 3936 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7349, - "estimated_input_tokens": 1838, - "estimated_eval_tokens": 3932 + "serialized_chars": 7358, + "estimated_input_tokens": 1840, + "estimated_eval_tokens": 3936 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly low-information spammy or off-topic issues with only superficial title similarity; none share a concrete underlying bug or actionable request.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44572 is the earliest of the exact \"\" reports and is the clearest representative of the repeated spam-like items, though the cluster as a whole does not form a true duplicate set.", + "summary": "These are not duplicates of one another; the cluster looks like unrelated spam/generic placeholder issues rather than a single concrete bug or request.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items are unrelated low-signal/spam posts with no shared concrete problem to unify around.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44572 is the best single representative for manual triage because it is an exact spam placeholder and matches another identical spam issue, but it is not a substantive canonical bug report.", + "best_issue_reason": "No issue is a good global representative because none describes a real, specific underlying bug or enhancement.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are placeholder spam titles with no shared concrete bug or change to deduplicate." + "reason": "Both are generic spam placeholders, but there is no evidence they describe the same underlying issue." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Same as above: identical spam-style placeholders, but no underlying actionable issue is described." + "reason": "Same generic spam title is not enough to conclude a shared bug or request." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is spam and the other is a generic documentation suggestion; they do not describe the same problem." + "reason": "One is spam and the other is a documentation suggestion; they are unrelated." }, { "left": "issue:44371", "right": "issue:47036", "accept": false, - "reason": "Spam placeholder versus a community benchmark tip; unrelated topics." + "reason": "Spam placeholder versus community benchmark tip; no common concrete issue." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder versus a 'Google LLC' message; no common bug/change." + "reason": "Spam placeholder versus a specific outreach-style title; not the same underlying problem." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Generic documentation suggestion and spam placeholder are not the same underlying report." + "reason": "Documentation suggestion and spam placeholder are different kinds of posts with no shared bug." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "Documentation request versus spam placeholder; no shared actionable issue." + "reason": "Documentation suggestion and spam placeholder are unrelated." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "These are both spam placeholders, but there is no concrete bug/change to merge beyond superficial duplication." + "reason": "Identical spam titles alone do not establish the same underlying issue." }, { "left": "issue:44572", @@ -21163,19 +21155,19 @@ "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder versus 'Check out \"Google LLC\"'; no shared underlying issue." + "reason": "Spam placeholder versus outreach-style title; no shared concrete issue." }, { "left": "issue:44573", "right": "issue:47036", "accept": false, - "reason": "Spam placeholder and benchmark tip do not refer to the same problem." + "reason": "Spam placeholder and benchmark tip are unrelated." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder versus 'Check out \"Google LLC\"'; unrelated." + "reason": "Spam placeholder versus outreach-style title; not a duplicate." } ] }, @@ -21205,12 +21197,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "e0dbeebf7431ee553681add5c8eee9bdf3b9e1fe13d4c6f10967c83eacd85aea", + "prepared_review_unit_hash": "067ad7d90c043812202e83ad80b475e398a0e1d95855d0159a3b1cea6bb78dfe", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -21218,7 +21210,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -21227,24 +21219,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are likely duplicates: they share the exact title and a soft-similarity signal, but the packet does not expose the full bodies, so this should be treated as a probable rather than definitive match. #45145 remains the best canonical issue because it was created first and already has an inbound reference.", - "confidence": 0.76, - "canonical_issue_reason": "#45145 is the earliest of the pair and already has an inbound reference, so it is the strongest canonical target for this likely duplicate issue.", + "summary": "The pair looks like a likely duplicate issue proposal: the titles are identical, the bodies are similarly sized, and there is a soft textual-similarity signal with no distinguishing metadata. #45145 is the better canonical issue because it was created earlier and already has the only inbound reference.", + "confidence": 0.86, + "canonical_issue_reason": "#45145 is earlier and has the only inbound reference, so it is the best canonical issue to keep for this likely duplicate pair.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best representative issue because it is the first-created copy of what appears to be the same report, and it has the stronger existing linkage.", + "best_issue_reason": "#45145 is the strongest representative issue: it appears to describe the same proposal as #45151 and has the slight advantage of earlier creation plus an inbound reference.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": true, - "reason": "Probable duplicate: the titles are identical and the soft-similarity signal supports a shared underlying report, but the packet does not include body text, so this should be treated conservatively rather than as a definitive duplicate." + "reason": "Likely the same underlying proposal: identical titles, similar body length, and a soft similarity signal. The evidence is suggestive rather than conclusive, but it is enough to treat them as duplicates." } ] }, "evaluator_result": { "accept": false, - "feedback": "The canonical choice is reasonable, but the duplicate verdict is overstated for the packet. The evidence shown only supports same title plus a soft-similarity signal; the bodies/descriptions are not included, so calling this a \"clear duplicate pair\" or \"near-identical\" is stronger than the packet justifies. Please downgrade to a more conservative, probabilistic assessment or cite direct body evidence if available." + "feedback": "The canonical choice is reasonably grounded (earlier creation and the only inbound reference), but the duplicate verdict is a bit too strong for the packet. The evidence shown is identical titles, similar body lengths, and a soft-similarity signal; it does not directly establish that the bodies are the same or that there is no meaningful difference. The confidence of 0.99 is also too high for soft evidence only. A more conservative phrasing would be that they likely duplicate the same proposal." }, "evaluator_used": true, "retried": true, @@ -21325,40 +21317,40 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "c6dd57992f6102195702a3672c9fd11dc315206bd0e19472d20155cc4d9f3f59", + "prepared_review_unit_hash": "f29669185cace2791c0c7b6837b31c2c0abff251a92bb319610e907adf350fa0", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1476, - "estimated_input_tokens": 369, - "estimated_eval_tokens": 994 + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1476, - "estimated_input_tokens": 369, - "estimated_eval_tokens": 994 + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are unrelated one-off name/date entries and do not describe the same bug or request, so they should not be clustered as duplicates.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "These two issues are unrelated empty posts with different personal names and dates; there is no evidence they describe the same bug, request, or change.", + "confidence": 0.99, + "canonical_issue_reason": "Issue #45762 is not meaningfully better than #45763; both are blank, non-technical reports, so there is no true canonical issue in this cluster.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No issue is a good representative because neither contains actionable problem details or duplicate content.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Different people and different dates; no shared problem statement or actionable overlap, so this is not the same underlying issue." + "reason": "Different names and birth dates, with no shared problem statement or code/change context; the similarity is only superficial." } ] }, @@ -21388,40 +21380,40 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "f49aca81abe9b21dbc4d20d213fcb47d93d09a4b1163b3928594c707b7858406", + "prepared_review_unit_hash": "a86ced27b18d47c2be66e2b72303541c71e79a30ec9dae59c75a9953c6eac662", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1481, - "estimated_input_tokens": 371, - "estimated_eval_tokens": 998 + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1481, - "estimated_input_tokens": 371, - "estimated_eval_tokens": 998 + "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 closed issues are unrelated personal-name/date entries with no shared subject, code context, or discussion indicating duplication.", + "summary": "The two issues are not duplicates; they look like unrelated placeholder/person-name entries with different names and dates, and there is no shared technical context.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "No canonical issue: the items do not describe the same bug or request, so there is no duplicate target to consolidate around.", "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good canonical choice because they describe different people and different dates, with no evidence of a shared underlying bug or task.", + "best_issue_reason": "Neither issue is a good canonical representative; if one must be chosen, 45764 is only marginally cleaner formatted, but it is not meaningfully more canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "The titles refer to different names and dates; there is no indication they describe the same problem or request." + "reason": "Different names and different dates with no shared problem statement or technical context; this is not the same underlying issue." } ] }, @@ -21451,12 +21443,12 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "e3407f466a4136c19d876441cf32df52d17ec8061111316ce3cf03cc2ab688b0", + "prepared_review_unit_hash": "e0b9647706b23ee76e0ba3a125723894cdbc8e037126f94f2b882ed0460e3ed8", "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 }, @@ -21464,7 +21456,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1542, + "serialized_chars": 1541, "estimated_input_tokens": 386, "estimated_eval_tokens": 1028 }, @@ -21473,18 +21465,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two open issues are not duplicates: one requests a new token-classification model head for Qwen3_5, while the other asks for FP8 training support. They target different code paths and different feature areas.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 46295 is the broader, more general feature request and has inbound attention, but it is still unrelated to 45810; there is no true duplicate pair here.", + "summary": "These two issues are unrelated feature requests: one asks for a new Qwen3_5 token-classification/value-model class, while the other requests FP8 training support. They should not be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good canonical duplicate for the other. If a single representative is needed, 46295 is the more general platform feature request, while 45810 is a specific model-head addition request.", + "best_issue_reason": "Issue 45810 is the more specific request and reads like a concrete model-class addition, so it is the better representative of this cluster even though it is not a duplicate of 46295.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different requests with different underlying changes: adding Qwen3_5ForTokenClassification versus enabling FP8 training. They do not describe the same bug or change, so this soft match should be rejected." + "reason": "Different underlying requests: model/task support for Qwen3_5 token classification vs training-time FP8 support. No shared code-path or same bug/change." } ] }, @@ -21524,88 +21516,88 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "66fbcbc1aba291238813a319d56d7fde3de200111a7b688493c84762a102d246", + "prepared_review_unit_hash": "9ad8a8fe6beddd30c0dc8898ce7a116041874e61aebd10de13bf2c797d37af9d", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5729, - "estimated_input_tokens": 1433, - "estimated_eval_tokens": 3122 + "serialized_chars": 5723, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5729, - "estimated_input_tokens": 1433, - "estimated_eval_tokens": 3122 + "serialized_chars": 5723, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster looks like two strong duplicate candidates, not proven exact duplicates: 45870/45871 and 45872/45873 have identical titles, matching body lengths, and soft-similarity signals. 45869 is only thematically related and does not look like a duplicate. No PRs are present.", - "confidence": 0.88, - "canonical_issue_reason": "Issue 45870 is the best canonical representative for the first likely duplicate pair because it is the earlier of the two identical-title posts and has a clean, generic example title. The cluster also contains a separate likely duplicate pair at 45872/45873.", + "summary": "This cluster splits into one unrelated dataset issue plus two likely duplicate pairs: 45870/45871 and 45872/45873. The evidence is strongest on matching titles and matching body lengths for each pair, so the duplicate calls should stay conservative rather than claiming full-body identity.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 45870 is a good canonical representative for the 45870/45871 pair: it shares the same title and body length as 45871 and has slightly more discussion activity. It should not be treated as a representative for the separate 45872/45873 pair or the unrelated dataset issue.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45870 is the strongest single representative in this set: it is part of a high-similarity duplicate candidate pair, appears earliest among the repeated example issues, and is more neutral than the unrelated dataset post.", + "best_issue_reason": "Issue 45870 is the best single issue to keep as a representative item because it anchors one of the likely duplicate pairs and has marginally more engagement than 45871. The cluster still contains a second separate likely-duplicate pair and an unrelated issue, so no single issue fully covers the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Same broad financial/DataForge theme, but the titles describe different artifacts: a dataset bundle versus an example prediction writeup. Not the same underlying issue." + "reason": "Different topics: a dataset bundle listing versus a DataForge example request. Similar finance wording is not enough to make them the same issue." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Same broad financial/DataForge theme, but the titles describe different artifacts: a dataset bundle versus an example prediction writeup. Not the same underlying issue." + "reason": "Different topics: dataset bundle listing versus example request; no evidence they describe the same underlying issue." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Strong duplicate candidate: identical title, identical body length, and soft-similarity evidence. This looks like the same issue content, though exact duplication is not proven from metadata alone." + "reason": "Likely duplicate pair: same title and matching body length, with no signs of a different scope." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Related only by the same project/domain. The titles point to different requests: an example prediction post versus a time-series/scraping service request." + "reason": "Related DataForge/financial wording, but the requests differ in scope: one is an example request and the other adds a scraping-service component." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Related only by the same project/domain. The titles point to different requests: an example prediction post versus a time-series/scraping service request." + "reason": "Same broad theme as 45870/45872, but not the same concrete request." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Related only by the same project/domain. The titles point to different requests: an example prediction post versus a time-series/scraping service request." + "reason": "Related wording only; these appear to be different requests rather than the same underlying issue." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Related only by the same project/domain. The titles point to different requests: an example prediction post versus a time-series/scraping service request." + "reason": "Related wording only; the example request and the scraping-service idea do not look identical in scope." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Strong duplicate candidate: identical title, identical body length, and soft-similarity evidence. This looks like the same issue content, though exact duplication is not proven from metadata alone." + "reason": "Likely duplicate pair: same title and matching body length, with no evidence of a scope difference." } ] }, "evaluator_result": { "accept": false, - "feedback": "The summary and duplicate judgments overstate the evidence. The packet only shows identical titles and matching body lengths for 45870/45871 and 45872/45873, plus soft-similarity flags; it does not prove exact duplicates. Please downgrade to conservative language such as \"likely duplicate candidates\" or \"strong duplicate candidates\". The 45869 non-duplicate assessment and the note that no PRs are present are grounded." + "feedback": "The cluster-level conclusion is mostly grounded, but the analyst overstates the evidence in several places. From the packet we can see identical titles and matching body lengths for 45870/45871 and 45872/45873, but not the full bodies, so calling them \"exact duplicates\" or saying they have the \"same text\" is stronger than the packet supports. The summary should stay conservative (e.g., \"likely duplicate pairs\" or \"same title and same body length\"). The 45869-vs-others non-match is reasonable, but the canonical and soft-edge rationales need to avoid claiming exact identity that is not directly evidenced." }, "evaluator_used": true, "retried": true, @@ -21632,7 +21624,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "b3f16c1a21292edba5efb2266b3d5e0fb39b59633cc7d89fc169f5bdbb7f60d2", + "prepared_review_unit_hash": "b2dcfe5b1b28d9472aab7cf4a5816ab0d82ac79ccb766b6d687d853c102dfca2", "component_budget": { "node_count": 2, "item_count": 2, @@ -21654,24 +21646,27 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are closely related static-cache/linear-attention failures, but they do not look like the same exact bug report: one describes shape corruption during `early_initialization`, the other a Qwen3.5 `KeyError` for `linear_attention`.", - "confidence": 0.69, - "canonical_issue_reason": "Issue 46439 is the better canonical issue because it states the broader underlying problem in static cache initialization affecting linear-attention layers and points to the likely root cause more directly.", + "summary": "The two issues appear related around static cache and `linear_attention`, but the packet only provides titles and a soft-similarity signal, so the exact relationship is unconfirmed.", + "confidence": 0.66, + "canonical_issue_reason": "Issue 46439 is the more general and diagnostic report, since it describes the broader static-cache/linear-attention failure in a way that can cover multiple symptoms.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46439 is the strongest representative for the cluster: it is broader, more root-cause oriented, and more likely to subsume model-specific crash reports like 46441.", + "best_issue_reason": "Issue 46439 is the best representative issue because it is the more mechanism-oriented and less model-specific of the two.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "Related area, but the failure modes differ (`RuntimeError` from shape corruption vs `KeyError: 'linear_attention'`), so this does not look like the same underlying bug with enough certainty to merge." + "reason": "The titles suggest a possible connection around static cache and `linear_attention`, but the packet does not establish that they are the same underlying bug or the same code path." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The linkage is plausible, but the analyst overstates the evidence. The packet only gives titles plus a soft-similarity signal; it does not establish that these are the same underlying bug or the same code path. Phrases like \"same static-cache/linear-attention failure mode\" and \"same code path\" are stronger than the packet supports. A more conservative summary would say the two issues appear related around static cache and linear_attention, but the exact relationship is unconfirmed." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -21695,7 +21690,7 @@ "soft_pairs": [ "issue:46551|issue:46552" ], - "prepared_review_unit_hash": "6f65edee3cfc55a7fe62eff46dc6e9859f44dd8a0b5962284d39dbe5550f03f7", + "prepared_review_unit_hash": "c36bf37905a96077b5383f306cba80ca8b285a50af744daaa2b24c7b190cc49f", "component_budget": { "node_count": 2, "item_count": 2, @@ -21717,24 +21712,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are near-exact duplicates describing the same Wav2Vec2CTCTokenizer v5 special-token regression; issue #46552 is the better canonical thread because it is open and already has inbound references.", + "summary": "These two issues are near-perfect duplicates: same title, same apparent body, and the pair is flagged as maximally similar. Issue #46552 is the better canonical thread because it is open and already has inbound references.", "confidence": 0.99, - "canonical_issue_reason": "#46552 is open and appears to be the active duplicate target, while #46551 is already closed; both titles/bodies describe the same bug.", + "canonical_issue_reason": "#46552 is open and already attracting references, so it is the better live tracker for this identical bug report.", "canonical_pr_reason": null, - "best_issue_reason": "#46552 is the best issue to keep as the main tracker since it is the active open report and has existing references.", + "best_issue_reason": "#46552 is the strongest issue to keep as the cluster representative: same report as #46551, but open and more likely to be the active discussion thread.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46551", "right": "issue:46552", "accept": true, - "reason": "The titles and body content are identical and describe the same tokenizer bug, so these are duplicate reports of the same underlying issue." + "reason": "Same title and same issue content; this is the same underlying bug report." } ] }, "evaluator_result": { "accept": true, - "feedback": "The assessment is well grounded in the packet: both issues have identical titles and body lengths, and #46552 is indeed open with inbound references while #46551 is closed. The duplicate/canonical-thread conclusion is conservative and supported by the evidence." + "feedback": "Grounded in the packet: the issues share the same title, the pair has maximal soft-similarity (jaccard 1.0), and #46552 is open with inbound references, making it the better canonical thread." }, "evaluator_used": true, "retried": false, @@ -21761,7 +21756,7 @@ "soft_pairs": [ "issue:46657|issue:46692" ], - "prepared_review_unit_hash": "a80627fb4b4cb9bd13ef9d98011a336b9a78f0fb04cbeb44b880ec74400efd9c", + "prepared_review_unit_hash": "12746c6913e133b54fe8d0cf7a1e4130a26bb1690acb19eeda9d944b4ed63c8f", "component_budget": { "node_count": 2, "item_count": 2, @@ -21783,18 +21778,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues describe different training failures: one is a tensor shape mismatch during object detection training, the other is an out-of-memory problem after several epochs in RT-DETR v2. They should not be clustered as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": "Issue #46657 is the more concrete bug report, with a specific tensor-size mismatch error in object detection training; #46692 describes a different failure mode (OOM) and is not the same underlying problem.", + "summary": "The two issues describe different failure modes during training: one is a tensor-size mismatch, the other is an out-of-memory condition after several epochs. They are not the same underlying bug, so the soft duplicate link should be rejected.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 46657 is the cleaner representative if one must be chosen, because it states a concrete training error with a clear symptom; 46692 is a different resource-exhaustion problem, not the same defect.", "canonical_pr_reason": null, - "best_issue_reason": "#46657 is the best issue candidate because it identifies a precise runtime mismatch and is more likely to map to a specific fix or regression than the broader OOM report in #46692.", + "best_issue_reason": "46657 is the best issue to anchor the cluster because it is a specific, actionable bug report; 46692 appears to be an unrelated training instability/OOM report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46657", "right": "issue:46692", "accept": false, - "reason": "These are different symptom classes: tensor shape mismatch versus memory exhaustion. They may occur in training, but there is no indication they stem from the same underlying bug or would be fixed by the same change." + "reason": "Different symptoms and likely different root causes: tensor shape mismatch versus OOM after epochs. Shared training context is too broad to treat as a duplicate." } ] }, @@ -21824,7 +21819,7 @@ "soft_pairs": [ "issue:46822|issue:46823" ], - "prepared_review_unit_hash": "1d69164f83d70a8dbe9a23efef2266ad77b44a4a13b61ff29f1d0ec3f639d658", + "prepared_review_unit_hash": "2b7b5876fa34e0e0053c802631ea2d2f148dd825ffbeebffd0e284af1f69504a", "component_budget": { "node_count": 2, "item_count": 2, @@ -21846,18 +21841,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both model-loading / dispatch issues, but they describe different failure modes in different code paths: one is ZeRO-3 zero.Init missing partitioning for a composite submodule causing OOM, the other is 4-bit load_in_4bit with device_map='auto' incorrectly placing modules on CPU and raising a ValueError. They should not be clustered as duplicates.", + "summary": "The two issues describe different bugs in different loading paths: ZeRO-3 zero.Init partitioning/OOM for a composite multimodal submodule versus 4-bit device_map/max_memory dispatching a model to CPU. They should not be merged as duplicates.", "confidence": 0.97, - "canonical_issue_reason": "Issue 46822 is the better canonical issue because it is more specific about the failing path (ZeRO-3 zero.Init, composite minimax_m3_vl language submodule, multi-GPU OOM) and appears to target a concrete partitioning bug.", + "canonical_issue_reason": "Issue 46822 is the better canonical anchor: it is open, specific, and describes a concrete zero.Init partitioning failure; issue 46823 is a separate CPU-dispatch regression.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46822 is the strongest representative of its own bug class due to its precise reproduction context and explicit symptom chain; 46823 is a separate auto-device-map / quantization dispatch problem.", + "best_issue_reason": "46822 is the strongest standalone issue report because it clearly identifies the affected model path and failure mode (OOM during multi-GPU load).", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46822", "right": "issue:46823", "accept": false, - "reason": "Different underlying problems: ZeRO-3 partitioning/OOM versus 4-bit auto device mapping to CPU with ValueError. Similar broad theme, but not the same bug or fix." + "reason": "Different subsystems and failure modes: ZeRO-3 composite submodule partitioning/OOM vs 4-bit auto device mapping with max_memory causing CPU placement and ValueError. Not the same underlying bug." } ] }, @@ -21940,280 +21935,292 @@ "issue:47037|issue:47182", "issue:47091|issue:47182" ], - "prepared_review_unit_hash": "bb984fcc3159f74846fe30127ae21e41c6b94a8de974fce00486e29f8536bd2c", + "prepared_review_unit_hash": "9372e65054c9a0d9b394042362aa8daf8eccfc1f91d0159d7f76515c68c1fcf9", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 22737, - "estimated_input_tokens": 5685, - "estimated_eval_tokens": 11626 + "serialized_chars": 22741, + "estimated_input_tokens": 5686, + "estimated_eval_tokens": 11628 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 22737, - "estimated_input_tokens": 5685, - "estimated_eval_tokens": 11626 + "serialized_chars": 22741, + "estimated_input_tokens": 5686, + "estimated_eval_tokens": 11628 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a run of near-identical daily triage issues for Serge integration failures, but each item is a separate dated report rather than the same underlying bug. I found no true duplicate pair to merge; the similarity is driven by the shared template and subsystem name.", - "confidence": 0.96, - "canonical_issue_reason": "issue:47182 is the best representative because it is the newest open triage thread and the only currently active item in the series.", + "summary": "These are separate daily \"serge\" integration-failure triage issues, not duplicates of one underlying bug. The high similarity comes from a shared template and date-stamped titles, not evidence of the same defect.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:47182 is the most suitable issue to keep as the live reference point, but only as a representative triage tracker\u2014not because the others are duplicates.", + "best_issue_reason": "issue:47182 is the latest open triage thread and the best representative of the current series, but it should not be treated as a duplicate of the earlier dated reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46906", "right": "issue:46940", "accept": false, - "reason": "Different dated triage reports (2026-06-26 vs 2026-06-27); same template, but not the same incident." + "reason": "Different dated triage reports; similarity is template-based, not the same underlying failure." }, { "left": "issue:46906", "right": "issue:46948", "accept": false, - "reason": "Separate daily triage entries for different dates, not one bug report." + "reason": "Different dated triage reports; no evidence they track one shared bug." }, { "left": "issue:46906", "right": "issue:46969", "accept": false, - "reason": "Distinct day-specific triage issue, despite very similar wording." + "reason": "Separate daily reports for different dates, not duplicates." }, { "left": "issue:46906", "right": "issue:46985", "accept": false, - "reason": "These are consecutive daily triage issues, not the same underlying failure." + "reason": "Same triage template, but each issue is a distinct day\u2019s report." }, { "left": "issue:46906", "right": "issue:47037", "accept": false, - "reason": "Different report date and likely different integration run; similarity is only templated." + "reason": "The only overlap is the recurring triage wording; these are distinct reports." }, { "left": "issue:46940", "right": "issue:46948", "accept": false, - "reason": "Two separate daily triage reports; no evidence of the same concrete bug." + "reason": "Different daily triage issues; no shared concrete defect is shown." }, { "left": "issue:46940", "right": "issue:46969", "accept": false, - "reason": "Same series of template-matched triage issues, but different day-specific reports." + "reason": "Each issue corresponds to a different date's integration triage." }, { "left": "issue:46940", "right": "issue:46985", "accept": false, - "reason": "Not duplicates; they track different dates in the triage sequence." + "reason": "Templated similarity only; not enough to call the same bug." }, { "left": "issue:46940", "right": "issue:47049", "accept": false, - "reason": "Different dated triage items, not one shared issue." + "reason": "Separate dated triage threads, not one mergeable duplicate." }, { "left": "issue:46948", "right": "issue:46969", "accept": false, - "reason": "Separate daily integration triage reports with the same structure only." + "reason": "Different day triage reports with no concrete shared failure details." }, { "left": "issue:46948", "right": "issue:46985", "accept": false, - "reason": "Distinct dated triage entries; no shared concrete failure is shown." + "reason": "These appear to be independent daily reports, not duplicates." }, { "left": "issue:46948", "right": "issue:46998", "accept": false, - "reason": "Different daily reports, so not mergeable duplicates." + "reason": "No evidence of the same underlying issue beyond the template." }, { "left": "issue:46969", "right": "issue:46985", "accept": false, - "reason": "Same recurring triage template, but different report day and likely different run." + "reason": "Adjacent daily triage issues, but still separate reports." }, { "left": "issue:46969", "right": "issue:47037", "accept": false, - "reason": "Not the same issue; these are day-separated triage threads." + "reason": "The date-stamped triage pattern indicates distinct occurrences." }, { "left": "issue:46969", "right": "issue:47091", "accept": false, - "reason": "Different dated integration-failure triage reports." - }, - { - "left": "issue:46985", - "right": "issue:46998", - "accept": false, - "reason": "Separate daily triage issues, not the same bug fix target." + "reason": "Different daily integration-failure triage threads." }, { "left": "issue:46985", "right": "issue:47037", "accept": false, - "reason": "Same broad triage stream, but different concrete report instances." + "reason": "Same subsystem and template, but not the same concrete problem." }, { "left": "issue:46985", "right": "issue:47049", "accept": false, - "reason": "Distinct day-specific reports; similarity is superficial." + "reason": "Separate dated triage reports; similarity is superficial." }, { - "left": "issue:46985", - "right": "issue:47091", - "accept": false, - "reason": "Not duplicates; they correspond to different dates in the triage run." - }, - { - "left": "issue:46985", - "right": "issue:47102", + "left": "issue:46998", + "right": "issue:47053", "accept": false, - "reason": "Different dated triage tickets, not a single underlying defect." + "reason": "Both are triage issues, but for different dates and likely different failures." }, { "left": "issue:46998", "right": "issue:47037", "accept": false, - "reason": "Separate daily integration-failure reports." + "reason": "No sign these are the same bug; just recurring triage notes." }, { "left": "issue:46998", "right": "issue:47049", "accept": false, - "reason": "Different day and likely different failing integration run." + "reason": "Different daily report entries, not a duplicate pair." }, { - "left": "issue:46998", - "right": "issue:47053", + "left": "issue:46985", + "right": "issue:46998", "accept": false, - "reason": "Same template, but these are distinct triage entries." + "reason": "Date-stamped triage issues; no shared underlying fix target." }, { "left": "issue:46998", "right": "issue:47091", "accept": false, - "reason": "No evidence these two reports describe the same exact failure." + "reason": "Same template, but distinct integration-failure reports." }, { - "left": "issue:46998", - "right": "issue:47120", + "left": "issue:47037", + "right": "issue:47049", "accept": false, - "reason": "Day-separated triage issues; do not collapse as duplicates." + "reason": "Separate daily triage issues rather than one defect." }, { "left": "issue:47037", - "right": "issue:47049", + "right": "issue:47091", "accept": false, - "reason": "Different dated triage reports in the same series." + "reason": "No evidence of one shared code-path problem." }, { "left": "issue:47037", "right": "issue:47053", "accept": false, - "reason": "Separate reports, not one underlying bug." + "reason": "These are recurring triage threads, not duplicates." }, { - "left": "issue:47037", + "left": "issue:47049", + "right": "issue:47053", + "accept": false, + "reason": "Different dates and no concrete bug linkage." + }, + { + "left": "issue:47049", "right": "issue:47091", "accept": false, - "reason": "Same subsystem and template, but different incident dates." + "reason": "Template similarity only; not the same issue." }, { - "left": "issue:47037", - "right": "issue:47102", + "left": "issue:47053", + "right": "issue:47091", "accept": false, - "reason": "Not the same concrete bug; these are distinct daily triage issues." + "reason": "Separate daily triage reports with no shared defect details." }, { - "left": "issue:47037", - "right": "issue:47182", + "left": "issue:47053", + "right": "issue:47102", "accept": false, - "reason": "Different dated triage threads, so they should not be merged." + "reason": "Not enough evidence that they describe the same failure." }, { - "left": "issue:47049", - "right": "issue:47053", + "left": "issue:47091", + "right": "issue:47102", "accept": false, - "reason": "Separate day-specific triage reports." + "reason": "Same recurring triage wording, but different reports." }, { - "left": "issue:47049", + "left": "issue:46985", "right": "issue:47091", "accept": false, - "reason": "No same-incident evidence; just matching triage wording." + "reason": "Different day triage threads; no concrete duplicate signal." }, { - "left": "issue:47049", + "left": "issue:47037", "right": "issue:47102", "accept": false, - "reason": "Different daily integration failure reports." + "reason": "Separate daily failure triage issues." }, { - "left": "issue:47053", - "right": "issue:47091", + "left": "issue:46985", + "right": "issue:47102", "accept": false, - "reason": "Distinct triage issues for different dates." + "reason": "These look like successive reports, not one bug." }, { - "left": "issue:47053", + "left": "issue:47049", "right": "issue:47102", "accept": false, - "reason": "Same template, but not the same bug report." + "reason": "Distinct dated triage issues with only template overlap." }, { - "left": "issue:47053", + "left": "issue:47120", + "right": "issue:47182", + "accept": false, + "reason": "Both are triage reports for different dates; not the same underlying change/bug." + }, + { + "left": "issue:47102", "right": "issue:47120", "accept": false, - "reason": "Different day and distinct report context." + "reason": "Different daily integration-failure triage tickets." }, { - "left": "issue:47091", - "right": "issue:47102", + "left": "issue:47053", + "right": "issue:47120", "accept": false, - "reason": "Separate daily triage threads, not duplicates." + "reason": "No evidence these are duplicate defect reports." }, { - "left": "issue:47091", - "right": "issue:47182", + "left": "issue:47049", + "right": "issue:47120", "accept": false, - "reason": "Different dated integration-failure triage items." + "reason": "Same series, but each issue tracks a separate day." }, { - "left": "issue:47102", + "left": "issue:46998", "right": "issue:47120", "accept": false, - "reason": "These appear to be consecutive daily triage reports, not one issue." + "reason": "Templated similarity only; not a mergeable duplicate." }, { "left": "issue:47102", "right": "issue:47182", "accept": false, - "reason": "Different report dates; similarity is only in the shared template." + "reason": "Latest two reports are still separate daily triage issues." }, { - "left": "issue:47120", + "left": "issue:47053", + "right": "issue:47182", + "accept": false, + "reason": "Different dates and no concrete shared failure described." + }, + { + "left": "issue:47037", + "right": "issue:47182", + "accept": false, + "reason": "These are individual daily triage threads, not duplicates." + }, + { + "left": "issue:47091", "right": "issue:47182", "accept": false, - "reason": "Both are triage reports for different days, so not the same underlying bug." + "reason": "No indication of the same underlying bug; just recurring triage." } ] }, @@ -22243,12 +22250,12 @@ "soft_pairs": [ "issue:46983|issue:46984" ], - "prepared_review_unit_hash": "38104e944d46bc6f776d39ba53dc70a5235504bed553489084ab03dbc0e44d21", + "prepared_review_unit_hash": "f93d402c5ca82535b57513beb931925017ee2a8e27d914b7738cdd00eb6e2f85", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1491, + "serialized_chars": 1492, "estimated_input_tokens": 373, "estimated_eval_tokens": 1002 }, @@ -22256,7 +22263,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1491, + "serialized_chars": 1492, "estimated_input_tokens": 373, "estimated_eval_tokens": 1002 }, @@ -22265,24 +22272,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are near-exact duplicates: same title, very high textual similarity, and no evidence of different scopes or separate bugs. One should be treated as the canonical tracker for RWKV7 support and the other closed as a duplicate.", - "confidence": 0.99, - "canonical_issue_reason": "Issue #46984 is the better canonical choice because it has slightly more discussion activity, suggesting it is the more active or consolidated thread for the same request.", + "summary": "The two issues appear to be duplicates: identical title, very similar bodies, and high text similarity indicate the same RWKV7 model support request.", + "confidence": 0.98, + "canonical_issue_reason": "Issue #46984 is the better canonical because it has slightly more discussion activity, suggesting it is the more active/complete thread for the same request.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #46984 is suitable as the primary issue for this cluster since it matches the duplicate request exactly and appears to be the more active discussion.", + "best_issue_reason": "Issue #46984 is the best representative issue: same underlying request as #46983, with a bit more discussion and no evidence of a separate scope.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46983", "right": "issue:46984", "accept": true, - "reason": "Same title, highly similar content, and no differentiating scope; these are the same underlying feature request for RWKV7 model support." + "reason": "Same title and highly similar issue bodies; both are clearly the same RWKV7 model support request." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is grounded in the packet: both issues have the same title, high soft-similarity, and issue #46984 has more discussion activity, so selecting it as the canonical thread is conservative and reasonable. The soft-edge acceptance is also supported by the available evidence." + "feedback": "The conclusion is grounded in the packet: both issues share the same title, have high textual similarity, and issue #46984 has slightly more discussion activity. The representative/canonical choice is conservative and supported by the available evidence." }, "evaluator_used": true, "retried": false, @@ -22309,7 +22316,7 @@ "soft_pairs": [ "issue:47007|issue:47015" ], - "prepared_review_unit_hash": "88b8422f7c428dc52cbc8e40c9da3dcf52c9a9d09e84836a71c71eb109979992", + "prepared_review_unit_hash": "596babec546b9ffe3bdab8ad44908df6089afedd17bd3f3d0e26f2d2264e4f08", "component_budget": { "node_count": 2, "item_count": 2, @@ -22331,18 +22338,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues are topically related to AI-driven vulnerability discovery, but they read like separate writeups/articles rather than the same underlying issue. The overlap is broad and thematic only.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 47015 is the broader, more general title and is the best single representative if one must be chosen, but the pair does not look like duplicates.", + "summary": "Both closed issues are broadly about AI/LLM-driven vulnerability discovery, but they read like different topic/proposal posts rather than the same underlying bug or change. With no shared targets, filenames, or concrete code-path evidence, I would not merge them as duplicates.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 47015 is the more direct, self-contained statement of the cybersecurity-vulnerability-discovery topic, so it is the better anchor if this cluster must be represented by a single issue.", "canonical_pr_reason": null, - "best_issue_reason": "47015 is the most umbrella-like issue in the pair, but neither issue is a strong canonical duplicate target.", + "best_issue_reason": "Issue 47015 is the clearest and most specific of the two titles, making it the better representative for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47007", "right": "issue:47015", "accept": false, - "reason": "Both discuss AI-based vulnerability discovery, but the titles point to different conceptual pieces: one is about lessons from advanced LLMs and the other about a novel transformer application. No explicit links, shared files, or shared targets indicate the same underlying change." + "reason": "Only soft textual similarity is present. The titles describe different framings of AI-assisted vulnerability discovery, not the same concrete bug, code path, or change, and there is no supporting issue/filename evidence." } ] }, @@ -22372,12 +22379,12 @@ "soft_pairs": [ "issue:47138|issue:47139" ], - "prepared_review_unit_hash": "b7c701896e194fbbcc8decb999ab46a687cd197dec727a119197b9de175eae58", + "prepared_review_unit_hash": "4b4722fe03cc24e217115fd910af081d25be22a1b84076886312a98de6dba4cd", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1591, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -22385,7 +22392,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1591, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -22394,28 +22401,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster contains two near-identical open issues describing the same bug; issue 47139 has slightly more inbound references and is the better canonical representative.", - "confidence": 0.99, - "canonical_issue_reason": "Issue 47139 is the stronger canonical choice because it has the same title/content as 47138 and already has more inbound references, suggesting it is the more visible reference point.", + "summary": "47139 is the better canonical issue because it shares the exact title with 47138 and has more inbound references, but the duplicate link itself is only weakly supported by the available metadata.", + "confidence": 0.89, + "canonical_issue_reason": "47139 has the same title as 47138 and already has more inbound references, so it is the better canonical target for consolidation.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 47139 is the best overall issue candidate: it matches the duplicate exactly and appears to be the more established thread for the same bug report.", + "best_issue_reason": "47139 is the best issue candidate because it matches the same reported problem and is the more established reference point.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47138", "right": "issue:47139", - "accept": true, - "reason": "Exact same title and apparently the same report; these are duplicates about the same underlying bug." + "accept": false, + "reason": "The titles are identical, but there is not enough additional linkage or distinguishing detail in the packet to confidently treat them as the same underlying report." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet: the two open issues have identical titles, identical body length, and a soft-similarity score of 1.0. Choosing issue 47139 as the canonical representative is also conservative because it has more inbound references (2 vs 0)." + "accept": false, + "feedback": "The packet supports that the two issues have identical titles and that 47139 has more inbound references, so picking 47139 as the canonical target is reasonable. However, the analyst overstates the evidence by calling this a \"clear duplicate\" / \"evidently the same report\" based only on title match and very limited metadata. The soft-edge verdict should stay more conservative unless there is stronger linking evidence." }, "evaluator_used": true, - "retried": false, - "accepted_nontrivial_soft_edge": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -22430,7 +22437,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-52", + "cluster_id": "cluster-40546-54", "nodes": [ "pull_request:43451", "pull_request:44517", @@ -22440,6 +22447,7 @@ "pull_request:46267", "pull_request:46376", "pull_request:46395", + "pull_request:46834", "pull_request:46836", "pull_request:46895", "pull_request:46903", @@ -22471,16 +22479,17 @@ "pull_request:47179", "pull_request:47181", "pull_request:47184", - "pull_request:47198", "pull_request:47202", "pull_request:47204", "pull_request:47207", "pull_request:47210", "pull_request:47212", "pull_request:47216", - "pull_request:47218" + "pull_request:47218", + "pull_request:47223" ], "soft_pairs": [ + "pull_request:46834|pull_request:47223", "pull_request:46903|pull_request:47090", "pull_request:46903|pull_request:46958", "pull_request:46954|pull_request:46958", @@ -22521,12 +22530,12 @@ "pull_request:44517|pull_request:46376", "pull_request:47204|pull_request:47207", "pull_request:44775|pull_request:47181", - "pull_request:47093|pull_request:47143", "pull_request:47077|pull_request:47210", + "pull_request:47093|pull_request:47143", "pull_request:47175|pull_request:47179", "pull_request:47093|pull_request:47163", - "pull_request:46266|pull_request:47216", "pull_request:47163|pull_request:47179", + "pull_request:46266|pull_request:47216", "pull_request:46836|pull_request:47218", "pull_request:46376|pull_request:47218", "pull_request:44775|pull_request:46836", @@ -22542,460 +22551,306 @@ "pull_request:47035|pull_request:47089", "pull_request:46895|pull_request:47207", "pull_request:44775|pull_request:46376", - "pull_request:43451|pull_request:44775", - "pull_request:47035|pull_request:47198", - "pull_request:46376|pull_request:47173", - "pull_request:47035|pull_request:47204", - "pull_request:46196|pull_request:46376", - "pull_request:47113|pull_request:47204", - "pull_request:46266|pull_request:46376", - "pull_request:47035|pull_request:47069", - "pull_request:47035|pull_request:47113", - "pull_request:47069|pull_request:47113", - "pull_request:46266|pull_request:47034" + "pull_request:43451|pull_request:44775" ], - "prepared_review_unit_hash": "22cd4e38a323271547321024c74a5558c724f2cb688e9dbfc2d2a15121402261", + "prepared_review_unit_hash": "e979a438321fe272482268239b95d0b9d061dc4068d12a98fc1b5514af4e3c9c", "component_budget": { - "node_count": 52, - "item_count": 52, - "soft_pair_count": 153, - "serialized_chars": 190750, - "estimated_input_tokens": 47688, - "estimated_eval_tokens": 95632 + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 72, - "serialized_chars": 135001, - "estimated_input_tokens": 33751, - "estimated_eval_tokens": 67758 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 134113, + "estimated_input_tokens": 33529, + "estimated_eval_tokens": 67314 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly unrelated PRs; the only clear duplicate subgroup is the CohereASR training-loss double-shift fixes. Most other soft edges are different models, different docs, or different infrastructure changes and should stay separate.", - "confidence": 0.93, + "summary": "This is a heterogeneous cluster: most PRs are unrelated model additions, docs, test updates, or infrastructure changes. The only clear duplicate subgroup is the CohereASR training-loss double-shift fixes, which all target the same loss path; most other soft edges share a subsystem but not the same concrete change.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47090 is the cleanest representative of the only real duplicate subgroup: it fixes the same CohereASR double-shift bug as the nearby CohereASR PRs, is narrowly scoped, and has the most direct issue-linked implementation.", + "canonical_pr_reason": "PR 47090 is the cleanest representative of the duplicate CohereASR loss bug: it fixes the exact training-loss double-shift in the target model, keeps the loss helper semantics intact, and is more focused than the broader or rougher variants.", "best_issue_reason": null, - "best_pr_reason": "If one PR must be chosen as the representative for this cluster, 47090 is the best fit because it is the most focused and clearly documented instance of the CohereASR loss-shift fix; the rest of the cluster is too heterogeneous to serve as a single canonical change.", + "best_pr_reason": "PR 47090 is the strongest global PR candidate because it is narrowly scoped, well explained, and fixes the actual CohereASR code path; the nearby alternatives are either broader multi-model changes or less precise implementations.", "soft_edge_verdicts": [ + { + "left": "pull_request:46834", + "right": "pull_request:47223", + "accept": false, + "reason": "Same model area and issue target, but these fix different bugs: a method-name typo in forward vs config attribute placeholders/tests." + }, { "left": "pull_request:46903", "right": "pull_request:47090", "accept": true, - "reason": "Same underlying CohereASR loss double-shift bug in the same code path; both change the loss call to stop shifting labels twice." + "reason": "Both fix the same CohereASR training-loss double-shift bug in the same code path; the implementation details differ, but the underlying change is the same." }, { "left": "pull_request:46903", "right": "pull_request:46958", "accept": true, - "reason": "Same CohereASR double-shift fix in the same files; 46958 is broader overall, but the CohereASR change is the same bug." + "reason": "Same CohereASR loss double-shift bug and same affected forward path; these are alternative implementations of the same fix." }, { "left": "pull_request:46954", "right": "pull_request:46958", "accept": false, - "reason": "Different changes: GlmASR Pytest-failure cleanup versus CohereASR/PPFormulaNet training-loss fixes." + "reason": "Unrelated changes in different models: Pytest-v9 fp32-module handling vs CohereASR loss shifting." }, { "left": "pull_request:46958", "right": "pull_request:46974", "accept": false, - "reason": "Unrelated code paths: ASR loss handling versus a MusicGen generate typo." + "reason": "Different models and problems: CohereASR training loss vs a MusicGen generate typo." }, { "left": "pull_request:47072", "right": "pull_request:47089", "accept": false, - "reason": "Different subsystems: chat template JSON-schema parsing versus AqlmConfig validation messages." + "reason": "Different utilities and bugs: chat-template enum parsing vs AqlmConfig type-error messages." }, { "left": "pull_request:47202", "right": "pull_request:47204", "accept": false, - "reason": "Both touch Docker tooling, but they solve different problems: removing a builder workflow versus adding exporter test dependencies." + "reason": "Both touch docker/CI, but one removes the old image builder while the other adds exporter dependencies to torch-light; not the same change." }, { "left": "pull_request:47077", "right": "pull_request:47078", "accept": false, - "reason": "Both are Japanese docs edits, but one is a translation patch and the other is wording/toc cleanup; not the same change." + "reason": "Both are Japanese docs edits, but one is a DistilBERT translation PR and the other is wording cleanup; not the same underlying change." }, { "left": "pull_request:47108", "right": "pull_request:47202", "accept": false, - "reason": "Unrelated infrastructure changes: dependabot-friendly doc-builder comments versus deleting the CI docker-image builder." + "reason": "Doc-builder comment tweak vs docker image builder deletion; unrelated changes." }, { "left": "pull_request:47077", "right": "pull_request:47212", "accept": false, - "reason": "Different docs topics: Japanese DistilBERT translation versus kv-cache documentation updates." + "reason": "Japanese docs translation vs kv-cache documentation updates; different content and purpose." }, { "left": "pull_request:47173", "right": "pull_request:47184", "accept": false, - "reason": "Different test fixes for different features; shared accelerator language is too broad to treat as the same change." + "reason": "Both are test updates for accelerator-related behavior, but they cover different models and different test expectations." }, { "left": "pull_request:46266", "right": "pull_request:46957", "accept": false, - "reason": "Different objectives: DINO conversion/modernization work versus finegrained FP8 decoding changes." + "reason": "Different code paths: DINO conversion/mapping updates vs generation-time FP8 kernel routing." }, { "left": "pull_request:47166", "right": "pull_request:47173", "accept": false, - "reason": "Different test suites and different issues: quantization output updates versus XPU expectation fixes." + "reason": "Unrelated test updates in different areas: quantization expectations vs XPU expectation fixes." }, { "left": "pull_request:44517", "right": "pull_request:47082", "accept": false, - "reason": "Different model additions: Qwen3-TTS versus RoMa keypoint matcher support." + "reason": "Different features entirely: Qwen3-TTS model/docs vs RoMa model support." }, { - "left": "pull_request:47034", - "right": "pull_request:47035", + "left": "pull_request:47035", + "right": "pull_request:47202", "accept": false, - "reason": "Related Neuron theme, but not the same bug/change: an experiment script update versus a torch-compile default backend tweak." + "reason": "Torch-compile Neuron defaulting and docker image cleanup are unrelated changes." }, { "left": "pull_request:46912", "right": "pull_request:47179", "accept": false, - "reason": "Different scope: RoPE test expansion versus Granite/GraniteMoE SWA model support." + "reason": "RoPE test improvements are unrelated to Granite-SWA model support." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47035", + "accept": false, + "reason": "Related Neuron work, but one adds a training example while the other changes the default compile backend; not the same change." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47173", + "accept": false, + "reason": "Both are accelerator/XPU-oriented tests, but they are for different models and different behaviors." }, { "left": "pull_request:46995", "right": "pull_request:47163", "accept": false, - "reason": "Different areas: flash-linear-attention availability versus paged-attention XPU test support." + "reason": "Flash-linear-attention availability change vs paged-attention XPU tests; different bugs." }, { "left": "pull_request:46196", "right": "pull_request:47210", "accept": false, - "reason": "Both concern attention, but they target different mechanisms: SDPA dispatch for audio models versus packed-row varlen attention routing." + "reason": "Both are attention-related, but one is broad SDPA dispatch support and the other is packed-row varlen attention routing." }, { "left": "pull_request:46376", "right": "pull_request:47202", "accept": false, - "reason": "Completely different changes: multimodal continuous batching versus docker-image builder removal." - }, - { - "left": "pull_request:47003", - "right": "pull_request:47204", - "accept": false, - "reason": "OpenVINO exporter work is unrelated to adding exporter test deps in torch-light Docker." + "reason": "Continuous-batching multimodal support is unrelated to docker image builder deletion." }, { "left": "pull_request:46395", "right": "pull_request:47181", "accept": false, - "reason": "Different model additions: Param2MoE versus Cosmos3 Edge." + "reason": "Different model additions: Param2MoE vs Cosmos3 Edge." }, { "left": "pull_request:47069", "right": "pull_request:47089", "accept": false, - "reason": "Unrelated fixes: nullable num_labels property versus AqlmConfig error-message wording." + "reason": "Num-labels None typing fix is unrelated to AqlmConfig error-message text." }, { "left": "pull_request:46196", "right": "pull_request:46995", "accept": false, - "reason": "Different attention-related work on different code paths and backends; not the same bug." + "reason": "Different attention features and code paths; not the same bug." }, { "left": "pull_request:47089", "right": "pull_request:47113", "accept": false, - "reason": "AqlmConfig validation messages versus audio feature/proc requirement annotations are unrelated." + "reason": "AqlmConfig validation wording is unrelated to audio feature/proc requirement decorators." }, { "left": "pull_request:46376", "right": "pull_request:46995", "accept": false, - "reason": "Different features: multimodal continuous batching versus flash-linear-attention backend availability." + "reason": "Continuous batching and flash-linear-attention availability are separate changes." }, { "left": "pull_request:46912", "right": "pull_request:47173", "accept": false, - "reason": "RoPE testing changes are unrelated to XPU expectation adjustments." + "reason": "RoPE test coverage vs XPU test expectation updates are unrelated." }, { "left": "pull_request:46196", "right": "pull_request:47202", "accept": false, - "reason": "Different subsystems entirely: audio SDPA support versus docker image cleanup." + "reason": "SDPA dispatch support and docker image cleanup do not share the same underlying change." }, { "left": "pull_request:46995", "right": "pull_request:47181", "accept": false, - "reason": "Flash-linear-attention availability and Cosmos3 Edge support are unrelated changes." + "reason": "Flash-linear-attention availability and Cosmos3 Edge model support are unrelated." }, { "left": "pull_request:47207", "right": "pull_request:47218", "accept": false, - "reason": "Different fixes: Swin2SR padding logic versus test image-loading refactor." + "reason": "Image padding logic fix vs test image-loading refactor; not the same bug." }, { "left": "pull_request:44775", "right": "pull_request:47077", "accept": false, - "reason": "Unrelated documentation changes: N-D parallelism docs versus Japanese DistilBERT translation." - }, - { - "left": "pull_request:47163", - "right": "pull_request:47179", - "accept": false, - "reason": "Different topics: XPU paged-attention tests versus Granite SWA model support." - }, - { - "left": "pull_request:46836", - "right": "pull_request:47218", - "accept": false, - "reason": "Unlimited OCR support and test image-loading refactor are unrelated." - }, - { - "left": "pull_request:46376", - "right": "pull_request:47218", - "accept": false, - "reason": "Continuous batching multimodal support versus test image-loading refactor are unrelated." - }, - { - "left": "pull_request:44775", - "right": "pull_request:46836", - "accept": false, - "reason": "Different docs and model-support changes; only shared documentation context." + "reason": "N-D parallelism docs and Japanese DistilBERT docs are unrelated documentation updates." }, { - "left": "pull_request:47108", + "left": "pull_request:46957", "right": "pull_request:47216", "accept": false, - "reason": "Doc-builder comment cleanup is unrelated to compressed-tensors FP8 support." + "reason": "FP8 kernel optimization vs compressed-tensors fp8 support; both quantization-adjacent, but different concrete changes." }, { "left": "pull_request:47082", "right": "pull_request:47181", "accept": false, - "reason": "Different model additions: RoMa versus Cosmos3 Edge." - }, - { - "left": "pull_request:46912", - "right": "pull_request:47082", - "accept": false, - "reason": "RoPE test work is unrelated to RoMa model support." + "reason": "RoMa keypoint matcher support and Cosmos3 Edge support are separate model additions." }, { - "left": "pull_request:47035", - "right": "pull_request:47089", - "accept": false, - "reason": "Torch-compile backend defaults and AqlmConfig error wording are unrelated." - }, - { - "left": "pull_request:46895", - "right": "pull_request:47207", - "accept": false, - "reason": "CohereASR loss-shift fix and Swin2SR padding fix are unrelated." - }, - { - "left": "pull_request:44775", - "right": "pull_request:46376", - "accept": false, - "reason": "Docs-only N-D parallelism update versus continuous batching feature work." - }, - { - "left": "pull_request:43451", - "right": "pull_request:44775", + "left": "pull_request:47175", + "right": "pull_request:47179", "accept": false, - "reason": "Molmo2 model support and N-D parallelism docs are unrelated." + "reason": "Olmo3 CI test expectation updates are unrelated to Granite-SWA model support." }, { "left": "pull_request:47093", "right": "pull_request:47143", "accept": false, - "reason": "Different model additions: OpenPangu-V2 versus NeuCodec." - }, - { - "left": "pull_request:47077", - "right": "pull_request:47210", - "accept": false, - "reason": "Japanese docs translation versus packed SDPA attention routing are unrelated." + "reason": "OpenPangu-V2 and NeuCodec are different model-addition PRs." }, { - "left": "pull_request:47093", - "right": "pull_request:47181", - "accept": false, - "reason": "OpenPangu-V2 and Cosmos3 Edge are different model-support PRs." - }, - { - "left": "pull_request:47175", + "left": "pull_request:47163", "right": "pull_request:47179", "accept": false, - "reason": "CI test expectation updates for OLMo3 versus Granite SWA model support are unrelated." + "reason": "Paged-attention XPU tests and Granite-SWA support are unrelated." }, { - "left": "pull_request:46957", + "left": "pull_request:46266", "right": "pull_request:47216", "accept": false, - "reason": "Finegrained FP8 decoding and compressed-tensors FP8 support are different quantization/export changes." - }, - { - "left": "pull_request:47082", - "right": "pull_request:47181", - "accept": false, - "reason": "RoMa keypoint matcher support and Cosmos3 Edge model support are unrelated." + "reason": "Modernizing Dino conversion/mapping and compressed-tensors fp8 support are different changes." }, { - "left": "pull_request:47166", + "left": "pull_request:46836", "right": "pull_request:47218", "accept": false, - "reason": "Quantization test updates versus image-loading test refactor are unrelated." - }, - { - "left": "pull_request:47035", - "right": "pull_request:47069", - "accept": false, - "reason": "Neuron torch-compile backend default and nullable num_labels are unrelated." - }, - { - "left": "pull_request:46895", - "right": "pull_request:47207", - "accept": false, - "reason": "CohereASR training-loss fix and Swin2SR padding fix are unrelated." - }, - { - "left": "pull_request:47093", - "right": "pull_request:47163", - "accept": false, - "reason": "Different model support versus paged-attention test support." - }, - { - "left": "pull_request:46266", - "right": "pull_request:47034", - "accept": false, - "reason": "DINO modernization work and Neuron SFT experiment scripts are unrelated." - }, - { - "left": "pull_request:47093", - "right": "pull_request:47181", - "accept": false, - "reason": "Different model-support PRs with no shared underlying bug or change." + "reason": "Unlimited OCR model support and image-loading test refactor are unrelated." }, { - "left": "pull_request:47179", - "right": "pull_request:47210", + "left": "pull_request:46376", + "right": "pull_request:47218", "accept": false, - "reason": "Granite SWA model support and packed SDPA attention routing are unrelated." + "reason": "Continuous batching multimodal support and test image loading refactor are unrelated." }, { - "left": "pull_request:46957", - "right": "pull_request:47202", + "left": "pull_request:44775", + "right": "pull_request:46836", "accept": false, - "reason": "Finegrained FP8 decoding and docker builder removal are unrelated." + "reason": "Docs for parallelism and Unlimited OCR model support are unrelated." }, { "left": "pull_request:47108", - "right": "pull_request:47181", - "accept": false, - "reason": "Doc-builder pin comment cleanup is unrelated to Cosmos3 Edge model support." - }, - { - "left": "pull_request:47077", "right": "pull_request:47216", "accept": false, - "reason": "Japanese documentation updates and compressed-tensors FP8 support are unrelated." + "reason": "Doc-builder pin comment cleanup is unrelated to compressed-tensors fp8 work." }, { - "left": "pull_request:47175", + "left": "pull_request:46957", "right": "pull_request:47216", "accept": false, - "reason": "OLMo3 CI expectation adjustments and compressed-tensors FP8 support are unrelated." + "reason": "Same broad quantization area, but different concrete changes and code paths." }, { "left": "pull_request:47082", - "right": "pull_request:47216", - "accept": false, - "reason": "RoMa model support and compressed-tensors FP8 support are unrelated." - }, - { - "left": "pull_request:47093", - "right": "pull_request:47216", + "right": "pull_request:47181", "accept": false, - "reason": "OpenPangu-V2 model support and compressed-tensors FP8 support are unrelated." + "reason": "Different model additions with no shared bug/change." }, { - "left": "pull_request:47163", - "right": "pull_request:47216", + "left": "pull_request:47166", + "right": "pull_request:47218", "accept": false, - "reason": "Paged-attention test support and compressed-tensors FP8 support are unrelated." + "reason": "Quantization test expectation updates are unrelated to image-loading refactors." }, { - "left": "pull_request:47034", - "right": "pull_request:47202", + "left": "pull_request:43451", + "right": "pull_request:44775", "accept": false, - "reason": "Neuron training experiment work and docker builder removal are unrelated." - }, - { - "left": "pull_request:46957", - "right": "pull_request:47181", - "accept": false, - "reason": "Quantization backend work and Cosmos3 Edge model support are unrelated." - }, - { - "left": "pull_request:47077", - "right": "pull_request:47179", - "accept": false, - "reason": "Japanese docs translation and Granite SWA model support are unrelated." - }, - { - "left": "pull_request:47166", - "right": "pull_request:47184", - "accept": false, - "reason": "Quantization test changes and accelerator test-skip changes are unrelated." - }, - { - "left": "pull_request:46995", - "right": "pull_request:47218", - "accept": false, - "reason": "Flash-linear-attention availability and image-loading test refactor are unrelated." - }, - { - "left": "pull_request:47072", - "right": "pull_request:47113", - "accept": false, - "reason": "Chat schema parsing and audio requirement cleanup are unrelated." - }, - { - "left": "pull_request:47108", - "right": "pull_request:47204", - "accept": false, - "reason": "Doc-builder comment cleanup and exporter-test Docker deps are unrelated." - }, - { - "left": "pull_request:47202", - "right": "pull_request:47216", - "accept": false, - "reason": "Docker builder removal and compressed-tensors FP8 support are unrelated." - }, - { - "left": "pull_request:47093", - "right": "pull_request:47218", - "accept": false, - "reason": "Model-support PR and test image-loading refactor are unrelated." + "reason": "Molmo2 model addition and N-D parallelism docs are unrelated." } ] }, @@ -23017,7 +22872,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-52", + "cluster_id": "cluster-40546-54", "nodes": [ "pull_request:40546", "pull_request:43451", @@ -23034,6 +22889,7 @@ "pull_request:46958", "pull_request:47003", "pull_request:47034", + "pull_request:47035", "pull_request:47069", "pull_request:47077", "pull_request:47078", @@ -23046,18 +22902,33 @@ "pull_request:47143", "pull_request:47163", "pull_request:47166", + "pull_request:47173", "pull_request:47175", "pull_request:47178", "pull_request:47181", "pull_request:47184", "pull_request:47198", + "pull_request:47204", "pull_request:47207", "pull_request:47210", "pull_request:47216", "pull_request:47218", - "pull_request:47219" + "pull_request:47219", + "pull_request:47223" ], "soft_pairs": [ + "pull_request:47035|pull_request:47198", + "pull_request:47035|pull_request:47204", + "pull_request:46376|pull_request:47173", + "pull_request:46196|pull_request:46376", + "pull_request:47113|pull_request:47204", + "pull_request:46196|pull_request:47223", + "pull_request:47089|pull_request:47223", + "pull_request:46266|pull_request:46376", + "pull_request:47035|pull_request:47113", + "pull_request:47035|pull_request:47069", + "pull_request:47069|pull_request:47113", + "pull_request:46266|pull_request:47034", "pull_request:47034|pull_request:47108", "pull_request:47069|pull_request:47082", "pull_request:46376|pull_request:47207", @@ -23069,20 +22940,21 @@ "pull_request:47113|pull_request:47175", "pull_request:47175|pull_request:47207", "pull_request:44517|pull_request:47181", + "pull_request:47113|pull_request:47223", "pull_request:43451|pull_request:46266", "pull_request:46376|pull_request:47198", "pull_request:43451|pull_request:44517", "pull_request:47077|pull_request:47216", "pull_request:47069|pull_request:47175", "pull_request:44517|pull_request:47143", + "pull_request:46895|pull_request:47223", "pull_request:46658|pull_request:46895", "pull_request:43451|pull_request:46957", "pull_request:46957|pull_request:47096", "pull_request:46376|pull_request:47184", - "pull_request:46957|pull_request:47003", "pull_request:47077|pull_request:47198", - "pull_request:43451|pull_request:47089", - "pull_request:43451|pull_request:46895", + "pull_request:46957|pull_request:47003", + "pull_request:47175|pull_request:47223", "pull_request:46895|pull_request:47090", "pull_request:46958|pull_request:47090", "pull_request:46895|pull_request:46958", @@ -23109,7 +22981,6 @@ "pull_request:46376|pull_request:47034", "pull_request:46658|pull_request:47143", "pull_request:47163|pull_request:47210", - "pull_request:46196|pull_request:47143", "pull_request:47143|pull_request:47163", "pull_request:47077|pull_request:47181", "pull_request:47034|pull_request:47216", @@ -23117,570 +22988,620 @@ "pull_request:47181|pull_request:47198", "pull_request:40546|pull_request:43451", "pull_request:43451|pull_request:47175", - "pull_request:47096|pull_request:47175", - "pull_request:47003|pull_request:47175", - "pull_request:43451|pull_request:46658", - "pull_request:46658|pull_request:47096", - "pull_request:46658|pull_request:47003", - "pull_request:43451|pull_request:47143", - "pull_request:43451|pull_request:47163", - "pull_request:43451|pull_request:47198", - "pull_request:43451|pull_request:47096", - "pull_request:43451|pull_request:47003", - "pull_request:47003|pull_request:47096", - "pull_request:43451|pull_request:47034", - "pull_request:47034|pull_request:47096" + "pull_request:47096|pull_request:47175" ], - "prepared_review_unit_hash": "50815708f205323d1ad6db151af675b21d75b236ef39186307b5454bc083e665", + "prepared_review_unit_hash": "d8b229da8b05a2988ceda914eba9faf4d7182ad4062c819c2d616530fab3c55c", "component_budget": { - "node_count": 52, - "item_count": 52, - "soft_pair_count": 153, - "serialized_chars": 190750, - "estimated_input_tokens": 47688, - "estimated_eval_tokens": 95632 + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 }, "budget": { - "node_count": 37, - "item_count": 37, + "node_count": 41, + "item_count": 41, "soft_pair_count": 72, - "serialized_chars": 120646, - "estimated_input_tokens": 30162, - "estimated_eval_tokens": 60580 + "serialized_chars": 127033, + "estimated_input_tokens": 31759, + "estimated_eval_tokens": 63774 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous: many are unrelated model adds, docs tweaks, or test-only changes. The only clear duplicate family is the CohereASR training-loss double-shift fix trio, plus one tightly related test refactor pair for COCO image loading.", + "summary": "This cluster is mostly a grab-bag of unrelated PRs (new model integrations, docs tweaks, test updates, exporter changes). Only two true duplicate mini-clusters stand out: the CohereASR training-loss double-shift fixes and the Aimv2 image-loading test migrations. Everything else is only broadly similar at the subsystem level, not the same concrete change.", "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:47090 is the strongest canonical PR anchor: it has the explicit issue target, covers the same CohereASR loss-shift code path as the earlier variants, and is the most complete expression of that fix.", + "canonical_pr_reason": "No single PR cleanly represents the whole mixed cluster; among the real duplicate subgroup, PR 47090 is the clearest canonical representative because it is the most explicit and complete CohereASR double-shift fix.", "best_issue_reason": null, - "best_pr_reason": "If a single PR must represent the cluster, 47090 is the best choice because it is the clearest true duplicate family in the set; the rest of the items are mostly unrelated changes.", + "best_pr_reason": "PR 47090 is the best global PR fit among the actual duplicates: it fixes a concrete, well-scoped code-path bug with an explicit issue tie-in, while most other items are unrelated or only superficially similar.", "soft_edge_verdicts": [ + { + "left": "pull_request:47035", + "right": "pull_request:47198", + "accept": false, + "reason": "Both touch backend/model plumbing, but they fix different features and code paths." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47204", + "accept": false, + "reason": "Neuron compile default vs exporter dependencies are unrelated changes." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47173", + "accept": false, + "reason": "Both are test/infrastructure oriented, but they address different subsystems and failures." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46376", + "accept": false, + "reason": "Both are generation-related, but one is SDPA dispatch and the other is continuous batching support." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47204", + "accept": false, + "reason": "Audio requirement fixes vs docker exporter dependencies are not the same change." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47223", + "accept": false, + "reason": "Unrelated generation/backend work; no shared concrete bug or change." + }, + { + "left": "pull_request:47089", + "right": "pull_request:47223", + "accept": false, + "reason": "Quantization config type-check messages vs TimesFM AttributeError fix are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46376", + "accept": false, + "reason": "Different modernization tasks in different areas; only broad similarity." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47113", + "accept": false, + "reason": "Neuron compile default and audio feature requirements are unrelated." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47069", + "accept": false, + "reason": "Trainium compile backend vs num_labels typing fix are not the same bug/change." + }, + { + "left": "pull_request:47069", + "right": "pull_request:47113", + "accept": false, + "reason": "Configuration typing and audio requirements are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47034", + "accept": false, + "reason": "Both are Trainium/backend adjacent, but one is a model-support PR and the other is an experiment script." + }, { "left": "pull_request:47034", "right": "pull_request:47108", "accept": false, - "reason": "Neuron SFT experiment code vs a doc-builder pin comment; unrelated changes." + "reason": "Trainium experiment code and doc-builder pin comments do not share the same underlying change." }, { "left": "pull_request:47069", "right": "pull_request:47082", "accept": false, - "reason": "One is a num_labels typing/None behavior fix, the other adds a new model; unrelated." + "reason": "A config property typing fix and a new keypoint matcher are unrelated." }, { "left": "pull_request:46376", "right": "pull_request:47207", "accept": false, - "reason": "Continuous batching multimodal support vs Swin2SR padding fix; different subsystems and bugs." + "reason": "Continuous batching multimodal support and Swin2SR padding fix are unrelated." }, { "left": "pull_request:47003", "right": "pull_request:47178", "accept": false, - "reason": "OpenVINO exporter work vs SigLIP checkpoint path cleanup; not the same concrete change." + "reason": "OpenVINO exporter support and SigLIP path fixes are different code paths." }, { "left": "pull_request:43451", "right": "pull_request:47078", "accept": false, - "reason": "Molmo2 model support vs Japanese doc wording; unrelated." + "reason": "Both are docs-heavy, but one adds a model and the other tweaks Japanese wording." }, { "left": "pull_request:47078", "right": "pull_request:47096", "accept": false, - "reason": "Japanese docs wording vs exportability work for Kimi; unrelated." + "reason": "Japanese docs wording/model text changes are unrelated to Kimi exportability." }, { "left": "pull_request:46912", "right": "pull_request:47113", "accept": false, - "reason": "RoPE test updates vs audio feature/proc requirement annotations; unrelated." + "reason": "RoPE test updates and audio feature/proc requirement fixes are unrelated." }, { "left": "pull_request:47003", "right": "pull_request:47078", "accept": false, - "reason": "OpenVINO exporter vs Japanese docs wording; no shared underlying bug or change." + "reason": "Exporter support and Japanese doc wording updates are not the same change." }, { "left": "pull_request:47113", "right": "pull_request:47175", "accept": false, - "reason": "Audio requirements fix vs Olmo3 CI expectation tweak; unrelated test maintenance." + "reason": "Audio requirements fixes and Olmo3 CI expectation updates are unrelated." }, { "left": "pull_request:47175", "right": "pull_request:47207", "accept": false, - "reason": "Olmo3 CI value updates vs Swin2SR over-padding fix; different issues." + "reason": "Different test/model areas with no shared bug or patch." }, { "left": "pull_request:44517", "right": "pull_request:47181", "accept": false, - "reason": "Qwen3-TTS support vs Cosmos3 Edge support; both are model additions, but not the same model or code path." + "reason": "Both add model docs/support, but for different models and different code paths." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47223", + "accept": false, + "reason": "Audio requirements and TimesFM window-size fix are unrelated." }, { "left": "pull_request:43451", "right": "pull_request:46266", "accept": false, - "reason": "Molmo2 support vs DINO conversion/modernization; unrelated model families." + "reason": "Both are model-integration PRs, but they add different models and code." }, { "left": "pull_request:46376", "right": "pull_request:47198", "accept": false, - "reason": "Continuous batching multimodal support vs GPTBigCode/GPTNeoX backend fixes for vLLM; separate code paths." + "reason": "Continuous batching and backend compatibility fixes are different changes." }, { "left": "pull_request:43451", "right": "pull_request:44517", "accept": false, - "reason": "Molmo2 and Qwen3-TTS are distinct model integrations; not duplicates." + "reason": "Different model additions; only the broad pattern is shared." }, { "left": "pull_request:47077", "right": "pull_request:47216", "accept": false, - "reason": "Japanese doc translation vs compressed-tensors fp8 support; unrelated." + "reason": "Japanese docs translation work and compressed-tensors quantization docs/features are unrelated." }, { "left": "pull_request:47069", "right": "pull_request:47175", "accept": false, - "reason": "num_labels None behavior fix vs Olmo3 CI test updates; unrelated." + "reason": "A config property typing fix and CI expectation refresh are unrelated." }, { "left": "pull_request:44517", "right": "pull_request:47143", "accept": false, - "reason": "Qwen3-TTS vs NeuCodec model support; both add models, but not the same change." - }, - { - "left": "pull_request:46658", - "right": "pull_request:46895", - "accept": false, - "reason": "Step-3.7 Flash work vs CohereASR loss-shift fix; unrelated." - }, - { - "left": "pull_request:43451", - "right": "pull_request:46957", - "accept": false, - "reason": "Molmo2 model addition vs finegrained FP8 v4 kernel/export changes; unrelated." - }, - { - "left": "pull_request:46957", - "right": "pull_request:47096", - "accept": false, - "reason": "Finegrained FP8 v4 kernel/export changes vs Kimi exportability work; related only broadly to export, not the same concrete fix." - }, - { - "left": "pull_request:46376", - "right": "pull_request:47184", - "accept": false, - "reason": "Continuous batching multimodal support vs test-annotation changes for Qwen3-Next/VideOMT; unrelated." - }, - { - "left": "pull_request:46957", - "right": "pull_request:47003", - "accept": false, - "reason": "Finegrained FP8 v4 vs OpenVINO exporter; different exporter/quantization work." - }, - { - "left": "pull_request:47077", - "right": "pull_request:47198", - "accept": false, - "reason": "Doc translation vs backend/vLLM fixes; unrelated." - }, - { - "left": "pull_request:43451", - "right": "pull_request:47089", - "accept": false, - "reason": "Molmo2 support vs AqlmConfig error-message cleanup; unrelated." - }, - { - "left": "pull_request:43451", - "right": "pull_request:46895", - "accept": false, - "reason": "Molmo2 model support vs CohereASR training-loss fix; unrelated." + "reason": "Different model-support additions with no shared concrete implementation." }, { "left": "pull_request:46895", "right": "pull_request:47090", "accept": true, - "reason": "Same CohereASR double-shift loss bug in the same files; 47090 is a refined version of 46895." + "reason": "Same CohereASR training-loss double-shift bug; same files and same fix direction." }, { "left": "pull_request:46958", "right": "pull_request:47090", "accept": true, - "reason": "Same CohereASR loss-shift bug and same code path; these are duplicate fixes for the same issue." + "reason": "Same CohereASR double-shift loss bug; 47090 is the cleaner follow-up fix for the same code path." }, { "left": "pull_request:46895", "right": "pull_request:46958", "accept": true, - "reason": "Both patch the same CohereASR training-loss double-shift problem in the same model code." + "reason": "Both target the same CohereASR loss double-shift issue in the same files." }, { "left": "pull_request:47218", "right": "pull_request:47219", "accept": true, - "reason": "Both refactor the same COCO image-loading test flow to use load_test_image and address the same integration failures." + "reason": "Both migrate the same image-loading tests to HF Hub fixtures/helper loading and appear to be the same underlying test change." }, { "left": "pull_request:47166", "right": "pull_request:47184", "accept": false, - "reason": "Both are test updates, but for different models and different failures; not the same underlying change." + "reason": "Both are test updates, but one is quantization coverage and the other is accelerator/device-map test adjustments." }, { "left": "pull_request:46912", "right": "pull_request:47175", "accept": false, - "reason": "RoPE tests vs Olmo3 CI updates; unrelated." + "reason": "Different fixes: RoPE test coverage vs Olmo3 CI expectation updates." }, { "left": "pull_request:46658", "right": "pull_request:46836", "accept": false, - "reason": "Step-3.7 Flash WIP vs Unlimited OCR model support; unrelated." + "reason": "Step3.7 flash work and Unlimited OCR are unrelated model additions." }, { "left": "pull_request:47163", "right": "pull_request:47166", "accept": false, - "reason": "Paged-attention XPU tests vs quantization test expectation changes; unrelated." + "reason": "Paged-attention XPU test support and quantization test updates are not the same underlying change." }, { "left": "pull_request:47163", "right": "pull_request:47184", "accept": false, - "reason": "Paged-attention XPU test support vs Qwen3-Next/VideOMT accelerator test edits; unrelated." + "reason": "Different test suites and different hardware-targeted adjustments." }, { "left": "pull_request:46509", "right": "pull_request:47143", "accept": false, - "reason": "Nemotron Omni support vs NeuCodec support; distinct model integrations." + "reason": "Nemotron Omni support and NeuCodec support are different model integrations." }, { "left": "pull_request:46509", "right": "pull_request:47181", "accept": false, - "reason": "Nemotron Omni support vs Cosmos3 Edge support; different models and code paths." + "reason": "Different model families and different integration code." }, { "left": "pull_request:46376", "right": "pull_request:46509", "accept": false, - "reason": "Continuous batching multimodal support vs Nemotron Omni model support; unrelated." + "reason": "Continuous batching changes and Nemotron Omni model support are unrelated." }, { "left": "pull_request:43451", "right": "pull_request:46836", "accept": false, - "reason": "Molmo2 vs Unlimited OCR; both model additions, but not the same model or fix." + "reason": "Different model additions with different subsystems and files." }, { - "left": "pull_request:46196", - "right": "pull_request:47143", + "left": "pull_request:46376", + "right": "pull_request:47181", "accept": false, - "reason": "SDPA dispatch support for wav2vec2-related models vs NeuCodec model support; unrelated." + "reason": "Continuous batching and Cosmos3 Edge model support are unrelated." }, { - "left": "pull_request:46836", - "right": "pull_request:47143", + "left": "pull_request:46509", + "right": "pull_request:46836", "accept": false, - "reason": "Unlimited OCR vs NeuCodec; separate model additions." + "reason": "Nemotron Omni and Unlimited OCR are unrelated model additions." }, { "left": "pull_request:46836", - "right": "pull_request:47163", + "right": "pull_request:47143", "accept": false, - "reason": "Unlimited OCR model support vs paged-attention XPU tests; unrelated." + "reason": "Unlimited OCR support and NeuCodec support are different model integrations." }, { "left": "pull_request:47143", "right": "pull_request:47181", "accept": false, - "reason": "NeuCodec vs Cosmos3 Edge are different model integrations." + "reason": "Different model additions; same broad pattern, not the same change." }, { "left": "pull_request:46376", - "right": "pull_request:47143", + "right": "pull_request:47163", "accept": false, - "reason": "Continuous batching multimodal support vs NeuCodec support; not the same code-path problem." + "reason": "Continuous batching internals vs paged-attention test coverage are different changes." }, { "left": "pull_request:43451", "right": "pull_request:46509", "accept": false, - "reason": "Molmo2 and Nemotron Omni are unrelated model additions." + "reason": "Different model-support PRs for different architectures." }, { "left": "pull_request:47163", "right": "pull_request:47181", "accept": false, - "reason": "Paged-attention XPU test support vs Cosmos3 Edge model support; unrelated." - }, - { - "left": "pull_request:46376", - "right": "pull_request:47163", - "accept": false, - "reason": "Continuous batching multimodal support vs paged-attention XPU tests; different areas." + "reason": "Paged-attention tests and Cosmos3 Edge model support are unrelated." }, { "left": "pull_request:46376", "right": "pull_request:47034", "accept": false, - "reason": "Continuous batching multimodal support vs Neuron SFT experiment code; unrelated." + "reason": "Continuous batching work and Trainium SFT experiment code are unrelated." }, { "left": "pull_request:46658", "right": "pull_request:47143", "accept": false, - "reason": "Step-3.7 Flash WIP vs NeuCodec model support; unrelated." + "reason": "Step3.7 flash tooling and NeuCodec model support are unrelated." }, { "left": "pull_request:47163", "right": "pull_request:47210", "accept": false, - "reason": "Paged-attention XPU tests vs sdpa varlen packed-attention routing; related only at a very high level, not the same concrete bug/change." + "reason": "Paged-attention XPU tests and packed SDPA varlen routing are different attention changes." }, { "left": "pull_request:47143", "right": "pull_request:47210", "accept": false, - "reason": "NeuCodec support vs packed SDPA routing; unrelated." + "reason": "NeuCodec support and packed SDPA routing are unrelated." }, { "left": "pull_request:47181", "right": "pull_request:47198", "accept": false, - "reason": "Cosmos3 Edge model support vs GPTBigCode/GPTNeoX vLLM backend fixes; unrelated." + "reason": "Different model-support work; no shared concrete bug or code path." }, { "left": "pull_request:40546", "right": "pull_request:43451", "accept": false, - "reason": "VibeVoice model support vs Molmo2 model support; both are model additions, but different models and code paths." + "reason": "Both are large model-support PRs, but they add different models and unrelated code." }, { "left": "pull_request:43451", "right": "pull_request:47175", "accept": false, - "reason": "Molmo2 support vs Olmo3 CI test update; unrelated." + "reason": "Molmo2 model support and Olmo3 CI expectations are unrelated." }, { "left": "pull_request:47096", "right": "pull_request:47175", "accept": false, - "reason": "Kimi exportability vs Olmo3 CI expectations; unrelated." - }, + "reason": "Exportability fixes for Kimi and Olmo3 CI updates 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": 1, + "component_count": 33, + "review_unit_index": 3, + "review_unit_count": 3, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-54", + "nodes": [ + "pull_request:43451", + "pull_request:46509", + "pull_request:46658", + "pull_request:47003", + "pull_request:47034", + "pull_request:47096", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47175", + "pull_request:47181", + "pull_request:47184", + "pull_request:47198", + "pull_request:47223" + ], + "soft_pairs": [ + "pull_request:47003|pull_request:47175", + "pull_request:43451|pull_request:46658", + "pull_request:46658|pull_request:47096", + "pull_request:46658|pull_request:47003", + "pull_request:43451|pull_request:47143", + "pull_request:43451|pull_request:47163", + "pull_request:43451|pull_request:47198", + "pull_request:43451|pull_request:47096", + "pull_request:43451|pull_request:47003", + "pull_request:47003|pull_request:47096", + "pull_request:43451|pull_request:47034", + "pull_request:47034|pull_request:47096", + "pull_request:43451|pull_request:47166", + "pull_request:47096|pull_request:47166", + "pull_request:43451|pull_request:47184", + "pull_request:43451|pull_request:47223", + "pull_request:47096|pull_request:47223", + "pull_request:43451|pull_request:47181", + "pull_request:47003|pull_request:47223", + "pull_request:47096|pull_request:47143", + "pull_request:47003|pull_request:47143", + "pull_request:47096|pull_request:47163", + "pull_request:47003|pull_request:47163", + "pull_request:46509|pull_request:47096" + ], + "prepared_review_unit_hash": "18548f9e86c9b6139b14daaf6430cfd514b2be221aee1eb91da233410030dd9a", + "component_budget": { + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 24, + "serialized_chars": 41303, + "estimated_input_tokens": 10326, + "estimated_eval_tokens": 20908 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are mostly unrelated: model additions, exporter work, test updates, and one backend/conversion fix. The shared similarity is broad Transformers plumbing, not a single underlying bug or change, so no soft edges should be merged.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true canonical PR emerges from this cluster; the items span unrelated features and tests. If one must be named, 43451 is the most substantial standalone model-add PR, but it is not a duplicate of the others.", + "best_issue_reason": null, + "best_pr_reason": "43451 is the strongest representative of the cluster only in the sense that it is the largest, most complete feature PR here; however, it should not be used as a duplicate sink for the others.", + "soft_edge_verdicts": [ { "left": "pull_request:47003", "right": "pull_request:47175", "accept": false, - "reason": "OpenVINO exporter vs Olmo3 CI test tweaks; unrelated." + "reason": "OpenVINO exporter work is unrelated to Olmo3 CI test expectation updates; different code paths and different goals." }, { "left": "pull_request:43451", "right": "pull_request:46658", "accept": false, - "reason": "Molmo2 model support vs Step-3.7 Flash build/conversion work; unrelated." + "reason": "Molmo2 model support and Step-3.7 flash checkpoint conversion are separate model integrations with different files, code paths, and release targets." }, { "left": "pull_request:46658", "right": "pull_request:47096", "accept": false, - "reason": "Step-3.7 Flash build/conversion work vs Kimi exportability; unrelated." + "reason": "Step-3.7 flash model conversion/loading is unrelated to Kimi exportability changes; they touch different models and mechanics." }, { "left": "pull_request:46658", "right": "pull_request:47003", "accept": false, - "reason": "Step-3.7 Flash build/conversion work vs OpenVINO exporter; unrelated." + "reason": "Step-3.7 flash checkpoint tooling is not the same change as adding an OpenVINO exporter." }, { "left": "pull_request:43451", "right": "pull_request:47143", "accept": false, - "reason": "Molmo2 vs NeuCodec are different model additions." + "reason": "Molmo2 support and NeuCodec support are distinct model additions with different configs, processors, and tests." }, { "left": "pull_request:43451", "right": "pull_request:47163", "accept": false, - "reason": "Molmo2 model support vs paged-attention XPU test additions; unrelated." + "reason": "Molmo2 model support is unrelated to adding XPU paged-attention tests for generation." }, { "left": "pull_request:43451", "right": "pull_request:47198", "accept": false, - "reason": "Molmo2 support vs GPTBigCode/GPTNeoX backend fixes; unrelated." + "reason": "Molmo2 model addition does not match the GPTBigCode/GPTNeoX backend/export compatibility fix." }, { "left": "pull_request:43451", "right": "pull_request:47096", "accept": false, - "reason": "Molmo2 model support vs Kimi exportability; unrelated." + "reason": "Molmo2 model support is unrelated to Kimi exportability and attention-backend changes." }, { "left": "pull_request:43451", "right": "pull_request:47003", "accept": false, - "reason": "Molmo2 support vs OpenVINO exporter; unrelated." + "reason": "Model addition work for Molmo2 is a different change than introducing OpenVINO exporter support." }, { "left": "pull_request:47003", "right": "pull_request:47096", "accept": false, - "reason": "OpenVINO exporter vs Kimi exportability; both touch export-related areas, but not the same concrete fix." + "reason": "OpenVINO exporter support and Kimi exportability are both exporter-related only at a very broad level; they fix different code paths and should not be merged as duplicates." }, { "left": "pull_request:43451", "right": "pull_request:47034", "accept": false, - "reason": "Molmo2 support vs Neuron SFT experiment; unrelated." + "reason": "Molmo2 model support is unrelated to the Neuron SFT experiment and training-script changes." }, { "left": "pull_request:47034", "right": "pull_request:47096", "accept": false, - "reason": "Neuron SFT experiment vs Kimi exportability; 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": 1, - "component_count": 33, - "review_unit_index": 3, - "review_unit_count": 3, - "status": "reviewed", - "reason": null, - "source": "llm", - "cache_hit": false, - "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-52", - "nodes": [ - "pull_request:43451", - "pull_request:46509", - "pull_request:47003", - "pull_request:47096", - "pull_request:47143", - "pull_request:47163", - "pull_request:47166", - "pull_request:47181", - "pull_request:47184" - ], - "soft_pairs": [ - "pull_request:43451|pull_request:47166", - "pull_request:47096|pull_request:47166", - "pull_request:43451|pull_request:47184", - "pull_request:43451|pull_request:47181", - "pull_request:47096|pull_request:47143", - "pull_request:47003|pull_request:47143", - "pull_request:47096|pull_request:47163", - "pull_request:47003|pull_request:47163", - "pull_request:46509|pull_request:47096" - ], - "prepared_review_unit_hash": "13cd4a86cd88e8d58db500300d1d35ca29646f629668554e97a9ef7c6c6547cb", - "component_budget": { - "node_count": 52, - "item_count": 52, - "soft_pair_count": 153, - "serialized_chars": 190750, - "estimated_input_tokens": 47688, - "estimated_eval_tokens": 95632 - }, - "budget": { - "node_count": 9, - "item_count": 9, - "soft_pair_count": 9, - "serialized_chars": 25014, - "estimated_input_tokens": 6254, - "estimated_eval_tokens": 12764 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": true, - "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated model-addition PRs, exporter work, and test-only changes. None of the soft edges look like the same underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the items are not duplicates and target different code paths, models, and test areas.", - "best_issue_reason": null, - "best_pr_reason": "No single PR is globally representative; the largest single change (e.g. Molmo2 or Cosmos3 Edge support) still does not cover the unrelated exporter and test-only PRs.", - "soft_edge_verdicts": [ + "reason": "Neuron finetuning experiment changes do not overlap with Kimi export/backend fixes." + }, { "left": "pull_request:43451", "right": "pull_request:47166", "accept": false, - "reason": "Molmo2 model support vs quantization test updates; different features and code paths." + "reason": "Molmo2 model support and quantization test expansions are different kinds of changes with no shared underlying bug." }, { "left": "pull_request:47096", "right": "pull_request:47166", "accept": false, - "reason": "Kimi exportability work vs GPTQ/MXFP4 test adjustments; unrelated changes." + "reason": "Kimi export/backend work is not the same as GPTQ/MXFP4 test adjustments." }, { "left": "pull_request:43451", "right": "pull_request:47184", "accept": false, - "reason": "Molmo2 model addition vs test decorator changes for Qwen3Next/VideomT; not the same underlying change." + "reason": "Molmo2 model support is unrelated to the Qwen3 Next multi-accelerator test annotation change." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47223", + "accept": false, + "reason": "Molmo2 model addition is unrelated to the TimesFM2.5 window_size AttributeError fix." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47223", + "accept": false, + "reason": "Kimi export/back-end changes and TimesFM2.5 config/test fixes address different models and failure modes." }, { "left": "pull_request:43451", "right": "pull_request:47181", "accept": false, - "reason": "Both add model support, but for different models (Molmo2 vs Cosmos3 Edge); not mergeable as one duplicate fix." + "reason": "Molmo2 and Cosmos3 Edge are separate model-support PRs with different architectures and files." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47223", + "accept": false, + "reason": "OpenVINO exporter support is unrelated to the TimesFM2.5 AttributeError fix." }, { "left": "pull_request:47096", "right": "pull_request:47143", "accept": false, - "reason": "Exportable Kimi vs NeuCodec model support; distinct features with different implementation targets." + "reason": "Kimi export/backend support and NeuCodec model support are unrelated changes." }, { "left": "pull_request:47003", "right": "pull_request:47143", "accept": false, - "reason": "OpenVINO exporter support vs NeuCodec model support; unrelated." + "reason": "OpenVINO exporter support does not describe the same underlying feature as NeuCodec model integration." }, { "left": "pull_request:47096", "right": "pull_request:47163", "accept": false, - "reason": "Kimi export work vs paged-attention XPU test additions; different concerns and code paths." + "reason": "Kimi export/backend fixes are unrelated to paged-attention test coverage for XPU." }, { "left": "pull_request:47003", "right": "pull_request:47163", "accept": false, - "reason": "OpenVINO exporter changes vs XPU paged-attention tests; unrelated." + "reason": "OpenVINO exporter work is not the same as adding XPU paged-attention tests." }, { "left": "pull_request:46509", "right": "pull_request:47096", "accept": false, - "reason": "Nemotron Omni support vs Kimi exportability; separate model families and separate change sets." + "reason": "Nemotron Omni support is a model-addition PR; Kimi exportability is a separate backend compatibility change." } ] }, @@ -23710,12 +23631,12 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "05d4ffe3fcad5148a561f475f5e3e8e8612f63e591d95472ad288498c43f16e4", + "prepared_review_unit_hash": "ff9566c255cab4a813d1ae96af9aa79ffb1502218d596d8be8f0b2054e7f3d6c", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4674, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -23723,7 +23644,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4674, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -23732,18 +23653,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests are unrelated feature additions in different models and code paths: TimesFM covariate forecasting vs. DINOv3 image classification support. They should not be clustered as duplicates.", - "confidence": 0.99, + "summary": "These PRs are unrelated feature additions in different model families: TimesFM covariate forecasting vs. DINOv3 image-classification support. They should not be clustered as duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #41224 is the cleaner representative of the cluster because it is a focused, single-purpose model capability addition with a narrow title and concentrated implementation surface.", + "canonical_pr_reason": "If one PR must represent the set, #41224 is the better canonical PR because it is a narrower, concrete feature addition with clearer model-specific scope and substantial review activity.", "best_issue_reason": null, - "best_pr_reason": "PR #41224 is the best standalone representative here since it describes one concrete change and is easier to identify than the broader TimesFM covariate feature work.", + "best_pr_reason": "#41224 is the strongest standalone PR in this cluster: it targets a specific code path, includes docs/tests, and has a well-defined feature boundary.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different models and different features: TimesFM covariate forecasting is not the same underlying change as DINOv3 image classification support, and they would not plausibly be merged into one PR." + "reason": "Different models and different changes: TimesFM forecasting with covariates vs. DINOv3 image classification support. No shared files, targets, or underlying bug/change." } ] }, @@ -23862,154 +23783,154 @@ "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "c0bb6d7e594d06bd661fb81f11039009f9b940be3c8c888ed06c304564933d71", + "prepared_review_unit_hash": "afdec1af5aa3c2511a2c414d5f7f685b48549a0aa87501e37592fc81e3586cd3", "component_budget": { "node_count": 12, "item_count": 12, "soft_pair_count": 17, - "serialized_chars": 31602, - "estimated_input_tokens": 7901, - "estimated_eval_tokens": 16058 + "serialized_chars": 31599, + "estimated_input_tokens": 7900, + "estimated_eval_tokens": 16056 }, "budget": { "node_count": 12, "item_count": 12, "soft_pair_count": 17, - "serialized_chars": 31602, - "estimated_input_tokens": 7901, - "estimated_eval_tokens": 16058 + "serialized_chars": 31599, + "estimated_input_tokens": 7900, + "estimated_eval_tokens": 16056 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly noise: the soft similarities are driven by broad subsystem overlap, not the same underlying change. The only true duplicate-like pair is the two `fixed_cross_entropy` PRs, which both patch the same loss helper in `src/transformers/loss/loss_utils.py` to forward extra `cross_entropy` kwargs.", + "summary": "The only likely duplicate pair is the two `fixed_cross_entropy` PRs; the rest are unrelated changes that merely touch adjacent transformer subsystems or share broad themes like model integration/testing.", "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is the clearest canonical representative: it has the explicit issue target, a concrete fix in `fixed_cross_entropy`, and more complete review/activity than its near-duplicate.", + "canonical_pr_reason": "PR 43251 is the best canonical choice because it directly targets the shared `fixed_cross_entropy` bug, explicitly links the tracking issue, and contains the more complete fix for forwarding supported `cross_entropy` kwargs.", "best_issue_reason": null, - "best_pr_reason": "PR 43251 is the best global representative because it is the most direct, fully-scoped implementation of the only real duplicate pair in the set; the other PRs are unrelated changes or only weakly similar by broad area.", + "best_pr_reason": "PR 43251 is the strongest representative of the cluster\u2019s only true duplicate change-set: same file, same function, same issue, and a superset-style implementation of the same fix.", "soft_edge_verdicts": [ { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Different problems: one adds missing tokenizer methods for Mistral, the other adds a dynamic-shape export test for Executorch. Same broad testing/utility theme, not the same fix." + "reason": "Different problems: Mistral tokenizer API gaps vs. Executorch dynamic-shape export testing." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Unrelated model work: Mistral tokenizer methods vs adding a new MobileLLM model. No shared bug or change path." + "reason": "Unrelated: tokenizer method additions for Mistral vs. adding a new MobileLLM model." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Both touch processor/model loading infrastructure, but one is about robust `from_pretrained` resolution and the other adds a new model package. Not the same change." + "reason": "Both are processor/model loading work, but they fix different models and different code paths." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss helper kwargs passthrough vs adding a new model. No shared code-path problem." + "reason": "`fixed_cross_entropy` kwargs forwarding is unrelated to adding MobileLLM model support." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Different areas: cross-entropy helper kwargs vs Mistral4 test/model fixes. Shared model code is too broad to be duplicate evidence." + "reason": "Cross-entropy loss plumbing is unrelated to Mistral4 model/test changes." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Both are test-related in different features, but they address unrelated behaviors and code paths." + "reason": "Dynamic export tests and Mistral4 fixes do not address the same underlying bug." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Executorch export test vs MobileLLM model addition are unrelated." + "reason": "Executorch export coverage vs. new model addition are different changes." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "Adding MobileLLM and fixing config resolution order are unrelated changes." + "reason": "New model onboarding and `cls.config` resolution ordering are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "New model addition vs Falcon output-interface refactor; no same underlying bug/change." + "reason": "MobileLLM introduction and Falcon output-interface refactoring are different code paths." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "New model addition vs model-creation skill/docs. The latter is meta tooling, not the same code fix." + "reason": "Model addition vs. repository skill/documentation changes; not the same change." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "These solve different problems in different layers; no shared concrete bug." + "reason": "Different subsystems and bug types; no shared concrete fix." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon output-capture refactor and Mistral4 test fixes are different model-specific changes." + "reason": "Falcon output-collection refactor is not the same as Mistral4 test/model fixes." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor vs model creation skill are unrelated." + "reason": "Falcon model refactor vs. model-creation skill PR are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test/model fixes vs CLIP-like `layrnorm` renaming are separate issues." + "reason": "Mistral4 test/model fixes and `layrnorm` typo renaming are different bugs with different affected files." }, { "left": "pull_request:44827", - "right": "pull_request:46048", + "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 fixes vs TDT loss kernel integration are unrelated model/loss changes." + "reason": "Mistral4 fixes are unrelated to the model-creation skill/large scaffold PR." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test fixes and config-resolution-order fix address different code paths." + "reason": "No common concrete bug: Mistral4 fixes vs. `cls.config` resolution order." }, { "left": "pull_request:44827", - "right": "pull_request:45213", + "right": "pull_request:46048", "accept": false, - "reason": "Model-specific fixes vs repository-wide model creation skill are not the same change." + "reason": "Mistral4 fixes are unrelated to the TDT loss kernel addition." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "New model implementation vs skill/docs tooling; no overlap in underlying bug or feature." + "reason": "Adding a model and adding a model-creation skill are not the same underlying change." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs change the same function (`fixed_cross_entropy`) to forward supported `nn.functional.cross_entropy` kwargs; they target the same concrete bug and could be merged into one PR." + "reason": "Same function, same file, same tracking issue, and both aim to forward supported `cross_entropy` kwargs; they could plausibly be merged into one fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall. The only duplicate-like pair is correctly identified as PRs 43251/43254, and the rest of the pairwise rejections stay conservative given the packet\u2019s mostly soft-similarity evidence. The canonical/best PR choice is also supported by the shared issue target and the stronger review/activity on 43251." + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted soft-edge pair, pull_request:43251 and pull_request:43254, is well-supported: same file, same function, same tracked issue, and the diffs are mergeable as one bugfix. The other rejected pairs are described at a high level without overstating evidence." }, "evaluator_used": true, "retried": false, @@ -24036,7 +23957,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "1d58ec6c382edb5cd3ac6a26fd56a6d69592722c9a3de1be6c1be15c9fcc125b", + "prepared_review_unit_hash": "ea38cc0f59f7a0f06fbe7daa85cb3a14c0a50ec80956fc5eff3beb475ade9b90", "component_budget": { "node_count": 2, "item_count": 2, @@ -24058,18 +23979,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are workflow dependency bumps, but they target different action families and different upgrade goals: PR 42975 is a broad Node 24 compatibility pass across many GitHub Actions, while PR 42976 specifically updates docker/build-push-action to newer versions. They are related, but not the same underlying change.", - "confidence": 0.93, + "summary": "Both pull requests are broad GitHub Actions maintenance bumps, but they target different underlying upgrades: one is a Node 24 compatibility sweep, the other is a latest-version bump focused on Docker/build actions. They overlap on some workflow files, but not on the same concrete change or fix.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the better representative of the cluster because it is the broader, more central workflow maintenance change, covering multiple actions for Node 24 compatibility rather than a narrower single-action bump.", + "canonical_pr_reason": "PR 42975 is the clearer representative: it has a specific compatibility goal (Node 24) and a coherent set of workflow-action updates across many files.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the strongest stand-in for the group since it spans many workflows and actions; PR 42976 is narrower and focused on docker build actions only.", + "best_pr_reason": "PR 42975 is the strongest standalone choice because its intent is explicit and it covers the broader workflow-action compatibility migration, whereas 42976 is a separate version bump for different actions.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "These are both GitHub Actions upgrade PRs, but they fix different upgrade targets and likely different compatibility concerns. PR 42975 updates checkout/cache/upload-artifact and other actions for Node 24 compatibility, while PR 42976 only bumps docker/build-push-action to v6. They are related maintenance changes, not the same concrete PR." + "reason": "Reject: both are workflow dependency updates, but they do not fix the same concrete problem. 42975 is about Node 24 compatibility for core GitHub actions; 42976 upgrades docker/build-push-action and related actions to newer versions. Shared workflow files alone are not enough to treat them as duplicates." } ] }, @@ -24104,12 +24025,12 @@ "pull_request:43448|pull_request:45490", "pull_request:45490|pull_request:46562" ], - "prepared_review_unit_hash": "1886349807c2fb4cf0712c5d157d6b4f3a444605b594259b8a7d0901b4f89459", + "prepared_review_unit_hash": "258393b9b1ae6f53dcd74e6d7b9f497c2273b498cd204e63896882c54f3cedce", "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 }, @@ -24117,7 +24038,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13046, + "serialized_chars": 13047, "estimated_input_tokens": 3262, "estimated_eval_tokens": 6780 }, @@ -24126,36 +24047,36 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are four separate model-addition pull requests (ViT NEPA, Molmo, CTSM, Step 3.7 Flash). They share the usual Transformers boilerplate and some auto-mapping/docs edits, but they target different model families and code paths, so they are not duplicates.", - "confidence": 0.97, + "summary": "All four items are independent model-addition PRs (ViT NEPA, Molmo, CTSM, Step3p7). Their overlap is limited to shared Hugging Face boilerplate like docs and auto-registration, not the same code-path fix or feature.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No single PR is a duplicate representative of the others; the cluster is composed of distinct model integrations rather than one underlying change.", + "canonical_pr_reason": "No canonical PR: these are distinct model integrations with different architectures, files, and test surfaces, so there is no duplicate target among them.", "best_issue_reason": null, - "best_pr_reason": "PR 46562 is the broadest and most complete of the set, with the most review activity and the most extensive model integration surface, but it still represents a different model addition rather than the same change as the others.", + "best_pr_reason": "Step3p7 is the strongest standalone representative in this set because it has the broadest integration surface and highest review activity, but it is still a separate model addition rather than a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different models and different implementation/code paths: ViT NEPA vs Molmo. Shared docs/auto-registry scaffolding is expected for model additions, not evidence of duplication." + "reason": "ViT NEPA and Molmo are different model integrations; shared registry/docs changes are boilerplate and not evidence of the same underlying change." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Different model families and distinct files beyond generic registry/docs updates; these are separate additions, not one bug fix or one merged PR." + "reason": "ViT NEPA vs CTSM: different model families and code paths. The overlap is only generic auto/docs wiring." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are unrelated model integrations. Overlap is limited to standard Transformers boilerplate like docs, auto mappings, and tests." + "reason": "Molmo and CTSM are separate new-model PRs; they do not appear to fix or add the same concrete implementation." }, { "left": "pull_request:45490", "right": "pull_request:46562", "accept": false, - "reason": "Both add new models, but they implement different architectures and touch different model-specific code paths. Similarity is only at the framework/integration level, so they should not be merged as duplicates." + "reason": "CTSM and Step3p7 both touch model-registration infrastructure, but they introduce different models and unrelated implementations, so they should not be merged as duplicates." } ] }, @@ -24204,12 +24125,12 @@ "pull_request:46539|pull_request:46922", "pull_request:46886|pull_request:46922" ], - "prepared_review_unit_hash": "c46867f4848d85fd338ddc49a32a7c8469ff3b91d504d877dfbc95a87f67fc2d", + "prepared_review_unit_hash": "df636795354635f36383c927b5ec64f4e2a6a4f0ff4628b871089c6d4641bf02", "component_budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19604, + "serialized_chars": 19601, "estimated_input_tokens": 4901, "estimated_eval_tokens": 10058 }, @@ -24217,7 +24138,7 @@ "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19604, + "serialized_chars": 19601, "estimated_input_tokens": 4901, "estimated_eval_tokens": 10058 }, @@ -24226,103 +24147,112 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are mostly unrelated despite generic titles and soft-similarity signals. The only close relationship is between the two YOLOS PRs, but they still address different fixes in the same model file rather than the same concrete bug.", - "confidence": 0.96, + "summary": "Only the two YOLOS PRs are close enough to treat as the same underlying fix; the rest are unrelated test-only or subsystem-specific changes despite similar titles/labels.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:46922 is the cleanest representative: it has an explicit issue target, a very small focused code change, and a clear runtime fix in YOLOS. The other PRs are broader, test-only, or infrastructure-oriented, and do not look like the same change.", + "canonical_pr_reason": "PR 46922 is the best representative because it addresses the concrete YOLOS device-mismatch path directly and also includes the model split guard, making it the more complete version of the 46886/46922 fix pair.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46922 is the strongest standalone PR in the set because it directly fixes a concrete YOLOS device-placement bug with minimal scope and explicit linkage to the tracked issue.", + "best_pr_reason": "PR 46922 is the strongest standalone PR: it targets the same YOLOS issue as 46886, fixes the runtime device placement bug in forward, and includes the supporting _no_split_modules change.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Both touch tests, but they are for different models and different failures; no shared code path or concrete bug." + "reason": "Both are generic test-fix PRs, but they touch different models and different code paths; the similarity is only in the vague title and test assertions." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper test cleanup vs FP8 dequant logic are unrelated changes in different subsystems." + "reason": "Whisper tokenizer/test cleanup and FP8 dequant logic are unrelated changes in different subsystems." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "Whisper tokenizer/test fix and generation watermarking validation are unrelated." + "reason": "Whisper decode tests and generation/watermarking validation are different features with no shared bug." }, { "left": "pull_request:43291", "right": "pull_request:46922", "accept": false, - "reason": "Whisper tests and YOLOS device-mismatch fixes are entirely different artifacts." + "reason": "Whisper test changes and YOLOS device-map fixes are unrelated model-specific issues." }, { "left": "pull_request:43665", "right": "pull_request:46539", "accept": false, - "reason": "These are test tweaks for CLIP/ViT versus a Phimoe CI/integration fix; not the same underlying bug." + "reason": "These are both test/integration PRs, but they concern different models and different failure modes." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Model test assertions are unrelated to FP8 loading/dequantization code." + "reason": "CLIP/ViT test edits do not overlap with the FP8 loading/dequant fix." }, { "left": "pull_request:43665", "right": "pull_request:46380", "accept": false, - "reason": "Different scope and subsystem: model tests versus generation configuration validation." + "reason": "Model test scaffolding changes are not the same as the generation watermarking validation fix." }, { "left": "pull_request:43665", "right": "pull_request:46922", "accept": false, - "reason": "Generic test fixes are not the same as a YOLOS runtime device-placement fix." + "reason": "Vague test fixes for CLIP/ViT are not the same underlying bug as YOLOS device placement." }, { "left": "pull_request:44815", "right": "pull_request:46539", "accept": false, - "reason": "FP8 quantization/dequantization changes do not match the Phimoe CI memory/device-map fix." + "reason": "FP8 dequantization/loading code and Phimoe CI integration setup are different problems." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "Different code paths: FP8 kernel/load path versus generation watermarking validation." + "reason": "These modify unrelated areas: fine-grained FP8 quantization versus generation config validation." }, { "left": "pull_request:44815", "right": "pull_request:46922", "accept": false, - "reason": "FP8 dequantization and YOLOS device handling are unrelated." + "reason": "No shared code path: FP8 kernel/loading changes versus YOLOS device-map behavior." }, { "left": "pull_request:46380", "right": "pull_request:46539", "accept": false, - "reason": "Generation watermarking validation and Phimoe integration CI are different problems." + "reason": "Generation watermarking validation and Phimoe CI memory/device setup are unrelated." }, { "left": "pull_request:46380", "right": "pull_request:46922", "accept": false, - "reason": "No shared bug or change: generation config validation versus YOLOS device mismatch." + "reason": "Different models and different bugs: generation config validation versus YOLOS device mismatch." }, { "left": "pull_request:46539", "right": "pull_request:46922", "accept": false, - "reason": "Phimoe CI/device-map environment changes and YOLOS model fixes are separate issues." + "reason": "Phimoe CI test-environment changes do not match YOLOS runtime device placement fixes." + }, + { + "left": "pull_request:46886", + "right": "pull_request:46922", + "accept": true, + "reason": "Both target the same YOLOS device_map=\"auto\" mismatch issue in the same file and would plausibly be merged into one PR; 46922 is the more complete version." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The 46886/46922 pair is well supported as the same YOLOS bugfix family: both target the same issue, touch the same file, and 46922 is a superset with the forward device-placement fix plus the _no_split_modules guard. The other soft-edge rejections also look appropriately cautious and not overstated." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -24348,52 +24278,52 @@ "pull_request:43488|pull_request:43532", "pull_request:43532|pull_request:46747" ], - "prepared_review_unit_hash": "a821db657a0ecb182516aa68053cb84305199a6b415ecad02fb759122af474d0", + "prepared_review_unit_hash": "25fc486194bee22a0ae809ed7a7b351a555757657573a307889a8bc27281686e", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6347, - "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": 6347, - "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 pull requests are unrelated \"don't merge\" test/CI changes: one tweaks packaging/versioning, one overhauls several comment/notification workflows and a test helper, and one changes a security-gate workflow trigger. They share only superficial repository/CI similarity, not the same underlying fix.", + "summary": "Three open \"don't merge\" PRs that are only superficially similar but modify different workflows/code paths and are not duplicates of one another.", "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No true duplicate PR exists here; if a representative anchor is needed, 46747 is the cleanest single-purpose workflow change, but it is not a duplicate of the others.", + "canonical_pr_reason": "No single PR cleanly represents the cluster: 43488 is a packaging/version-bump test PR, 43532 rewires several CI/commenting utilities, and 46747 changes a security-gate workflow trigger.", "best_issue_reason": null, - "best_pr_reason": "46747 is the most self-contained and narrowly scoped PR in the set, but it still addresses a different workflow behavior than the others.", + "best_pr_reason": "No best PR for dedupe purposes; none of these appear to fix the same underlying bug or change, so there is no safe canonical representative.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:46747", "accept": false, - "reason": "Different changes: 43488 edits setup/versioning and a test file, while 46747 only switches a security-gate workflow trigger. Same broad CI area, but not the same bug or fix." + "reason": "Both are test/don't-merge PRs, but they touch unrelated files and different behaviors: package metadata/dependency setup vs. a workflow trigger change. No shared concrete bug or mergeable code-path fix." }, { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "43488 is about packaging/version metadata and a Vit test tweak; 43532 changes multiple CI/workflow files plus testing utilities. They do not target the same concrete issue." + "reason": "Only a superficial overlap in a Vit test file; the actual changes are unrelated (setup/version/dependency changes vs. CI workflow and utility rewrites). Not the same underlying change." }, { "left": "pull_request:43532", "right": "pull_request:46747", "accept": false, - "reason": "Both touch workflows, but 43532 is a broad comment/notification CI refactor while 46747 only changes the PR security-gate trigger. Not mergeable as one duplicate fix." + "reason": "These modify different CI workflows and code areas, with no evidence they address the same bug or would be merged into one PR. Similarity is broad subsystem overlap only." } ] }, @@ -24423,7 +24353,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "230a1e24f77ec961e3d1f65f746ed67ba117ef9c640f3606f0e7c1fa5e45a217", + "prepared_review_unit_hash": "75ca5553ab98c59b121fedc23f4b97c2ac4216625144e734437d950c0421a984", "component_budget": { "node_count": 2, "item_count": 2, @@ -24445,18 +24375,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are related in theme but are distinct feature additions: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They share infrastructure files, but they target different pipelines, docs, and model integrations, so they should not be treated as duplicates.", - "confidence": 0.93, + "summary": "Two separate feature PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They overlap on shared documentation/registration scaffolding, but the actual pipelines, model integrations, and test coverage are different, so they are not duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single canonical PR for a duplicate cluster: the PRs implement different pipelines and touch different underlying model code paths.", "best_issue_reason": null, - "best_pr_reason": "PR 43613 is the broader, more comprehensive pipeline addition, covering multiple segmentation model integrations and a larger documentation/test surface. If one PR must represent the cluster, it is the stronger candidate, though it is not a duplicate of 43612.", + "best_pr_reason": "PR 43613 is the broader representative of the group, with more model integrations and test coverage, but it should not be treated as the same change as 43612.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different concrete additions: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation. They share supporting files, but the pipeline names, task docs, and model coverage differ, so they are not the same underlying change." + "reason": "Different features and code paths: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation with different model backends and tests. Shared docs/registration files are incidental scaffolding, not evidence of the same underlying change." } ] }, @@ -24486,12 +24416,12 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "f3f4f347c28f6f6d65c63f183fb7402bc345d7c2af3fe245e94b7bb5f03c519f", + "prepared_review_unit_hash": "7e89e940fbcbd9f05b9636714e2cd8f9baad3a4e48eac6bc24eb8b9ddf47ab1a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4292, + "serialized_chars": 4291, "estimated_input_tokens": 1073, "estimated_eval_tokens": 2402 }, @@ -24499,7 +24429,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4292, + "serialized_chars": 4291, "estimated_input_tokens": 1073, "estimated_eval_tokens": 2402 }, @@ -24508,18 +24438,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated trainer extensibility PRs in the same file: one changes gradient-accumulation loss scaling behavior, the other refactors signature-column inspection for custom trainers/PEFT models. They do not look mergeable into one change or the same underlying bug.", - "confidence": 0.93, + "summary": "The two PRs both modify `trainer.py` but address different trainer extension behaviors: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom/PEFT trainers. They are not the same underlying fix.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: the changes target separate code paths and solve different problems, so neither PR is a duplicate of the other.", "best_issue_reason": null, - "best_pr_reason": "PR #43651 is the stronger representative because it addresses a concrete training-step loss-scaling path with a narrowly scoped behavioral fix and explicit API hook.", + "best_pr_reason": "PR #43651 is the more concrete bugfix and easier representative of the cluster, but it is still not a duplicate of #43663.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Same file and broad trainer area, but different code paths and goals: loss scaling during training_step versus signature-column filtering. They are not the same bug/change and would not naturally merge into one PR." + "reason": "Different problems in the same file: loss scaling for gradient accumulation vs signature/column filtering. They would not plausibly be merged as one PR." } ] }, @@ -24549,12 +24479,12 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "007d04d0b0ef68bc32779c5516a19bc33da58622e6e10b46402186a879e842b6", + "prepared_review_unit_hash": "c30b4fc567f3b5071d1e63549bae4a7000b89b6f49dc459651f4cdde0560c577", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -24562,7 +24492,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -24571,18 +24501,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same CLI serve module and both mention a TypeAdapter NameError, but they appear to fix different runtime problems with different scopes. They are related, but not close enough to treat as the same change.", - "confidence": 0.83, + "summary": "Two open PRs touch the same CLI serve file and both address TypeAdapter/annotation-related NameErrors, but they appear to fix different concrete cases and use different mechanisms. I would not merge them as duplicates.", + "confidence": 0.73, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue is present in the cluster. As a representative PR, pull_request:43842 is the tighter fix: it explicitly targets the pydantic/TypeAdapter import problem and keeps the change narrowly scoped to conditional typing imports.", + "canonical_pr_reason": "PR 43656 is the broader, earlier fix with more discussion and a larger typing/import cleanup in the same code path, so it is the better canonical representative of this cluster.", "best_issue_reason": null, - "best_pr_reason": "pull_request:43842 is the better standalone representative because its title and diff clearly identify the concrete failure mode ('TypeAdapter NameError when pydantic is not installed') and the patch is narrowly focused on that path.", + "best_pr_reason": "PR 43656 is the strongest standalone PR here because it appears to address the core CLI annotation NameError more comprehensively, while 43842 is a narrower follow-up aimed at the optional pydantic-import case.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Do not merge: both are in the same file and reference TypeAdapter, but 43656 adds future-annotations/typing cleanup while 43842 specifically fixes runtime failure when pydantic is absent. They look like adjacent fixes to related code, not the same underlying bug or one mergeable PR." + "reason": "Both are in `src/transformers/cli/serve.py` and both mention TypeAdapter/annotations, but the diffs point to different fixes: 43656 is a broader future-annotations/typing cleanup, while 43842 specifically handles the `pydantic`-missing case by guarding the import and quoting the annotation. They do not look like the same concrete bug fix." } ] }, @@ -24775,12 +24705,12 @@ "pull_request:44024|pull_request:44076", "pull_request:44073|pull_request:44074" ], - "prepared_review_unit_hash": "592fb55459860b9fea7860eb22de9317d833ad084a76c2f8a5bce4272e07f42d", + "prepared_review_unit_hash": "f90230f90790d39131cc31d2133b3286ea6618ee22b71ca74f0fc5504ba70c5d", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 76, - "serialized_chars": 84119, + "serialized_chars": 84117, "estimated_input_tokens": 21030, "estimated_eval_tokens": 42316 }, @@ -24788,7 +24718,7 @@ "node_count": 28, "item_count": 28, "soft_pair_count": 72, - "serialized_chars": 76940, + "serialized_chars": 76938, "estimated_input_tokens": 19235, "estimated_eval_tokens": 38726 }, @@ -24797,348 +24727,372 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a heterogeneous cluster of model-specific output-tracing refactors. Most pairs are only thematically similar and should stay separate; the only clear duplicate-like subcluster is GPT-Neo (44018/44068) and the GPT-J/CodeGen family (44066/44085/44722).", - "confidence": 0.78, - "canonical_issue_reason": null, - "canonical_pr_reason": "No issue artifact is present, so the best canonical PR is 44722: it is the most complete and explicit issue-linked representative of the GPT-J/CodeGen output-tracing refactor subcluster.", - "best_issue_reason": null, - "best_pr_reason": "44722 is the strongest single representative of the cluster\u2019s main duplicateable subtheme because it explicitly targets the tracking issue and overlaps directly with the GPT-J/CodeGen changes in 44066 and 44085.", + "summary": "This cluster is mostly a set of model-specific refactors to standardize output tracing with capture_outputs/can_return_tuple. They share a common umbrella tracker (#43979), but most are not true duplicates because they touch different models and code paths. The only strong duplicate-like pairs are same-model revisions for GPT-Neo and GPT-J/CodeGen.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #43979 is the shared umbrella tracker referenced by many PRs and best matches the cluster\u2019s common output-tracing refactor theme.", + "canonical_pr_reason": "PR #44722 is the broadest and most representative PR: it covers the core GPT-J/CodeGen code paths, explicitly ties to #43979, and looks like the most complete revision of that change.", + "best_issue_reason": "No standalone issue artifact is present in the cluster; #43979 is still the closest canonical tracker for the shared refactor series.", + "best_pr_reason": "PR #44722 is the strongest standalone representative because it subsumes narrower GPT-J-only work and most clearly captures the shared change pattern.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families and different code paths; only the refactor pattern matches." + "reason": "Different model code paths (CVT/FNet vs GPT-J); same output-tracing theme is too broad." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "CVT/FNet output-tracing work is unrelated to DeBERTa's attention/output refactor." + "reason": "Different models and implementations; only the refactor pattern is shared." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Both are GPT-J output-tracing refactors on the same model file; 44066 just also includes CodeGen, so they could plausibly be merged into one PR." + "reason": "Same GPT-J output-tracing refactor on the same code path; the GPT-J-only PR looks like a narrower version of the GPT-J/CodeGen change." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "Both are CNN-style image model refactors, but they touch different models and concrete implementations." + "reason": "Different models (RegNet/ResNet vs EfficientNet); this is a shared refactor pattern, not the same bug/change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet vs GPT-J/CodeGen: same umbrella theme, but not the same underlying bug or code path." + "reason": "EfficientNet vs GPT-J/CodeGen are unrelated concrete changes despite the shared standardization theme." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "Different model families and unrelated implementation paths." + "reason": "Different model families and code paths; not the same underlying change." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTa are unrelated; this is only broad standardization similarity." + "reason": "MobileNetV2 and DeBERTa are unrelated model-specific refactors." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "Different models and code paths; same refactor style is not enough." + "reason": "GPT-J/CodeGen versus MGP-STR are different concrete code paths." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen versus MPT is a different concrete change set." + "reason": "GPT-J/CodeGen and MPT are separate model implementations; same refactor style is not enough." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo model, same file, same standardized output-tracing change; 44068 looks like a follow-up/rewrite of 44018." + "reason": "Same GPT-Neo model and same output-tracing refactor; this looks like a later revision of the same concrete change." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo and GPT-J are different code paths despite similar decorator migration." + "reason": "GPT-Neo versus GPT-J/CodeGen are different model paths." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen work is unrelated to GPT-Neo." + "reason": "Different models and files; only the standardized output-tracing theme is shared." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet and GPT-J/CodeGen are unrelated implementations." + "reason": "ResNet is unrelated to GPT-J/CodeGen; same umbrella refactor pattern only." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J/CodeGen are different models and changes." + "reason": "MGP-STR and GPT-J/CodeGen are different concrete code paths." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet versus GPT-J: same pattern, different concrete code path." + "reason": "ResNet and GPT-J are unrelated model-specific changes." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet are different model implementations." + "reason": "GPT-Neo and ResNet are different underlying changes." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet and MPT are unrelated." + "reason": "ResNet and MPT are unrelated model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert are different models with different tracing logic." + "reason": "GPT-Neo and Flaubert are different concrete code paths." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert are different code paths." + "reason": "GPT-Neo and Flaubert are unrelated model-specific refactors." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo is unrelated to GPT-J/CodeGen." + "reason": "GPT-Neo and GPT-J/CodeGen are distinct models and changes." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT and GPT-J/CodeGen are different refactors in different model code." + "reason": "MPT and GPT-J/CodeGen are different code paths." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo and GPT-J are different model families." + "reason": "GPT-Neo and GPT-J are different model implementations." }, { "left": "pull_request:44018", - "right": "pull_request:44059", + "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and GPT-2 are separate concrete implementations." + "reason": "GPT-Neo and MGP-STR are unrelated changes." }, { "left": "pull_request:44071", "right": "pull_request:44085", "accept": false, - "reason": "MPT versus GPT-J: only broad output-tracing similarity." + "reason": "MPT and GPT-J are different concrete code paths." }, { - "left": "pull_request:44086", - "right": "pull_request:44068", + "left": "pull_request:44068", + "right": "pull_request:44086", "accept": false, - "reason": "MGP-STR and GPT-Neo are unrelated model implementations." + "reason": "GPT-Neo and MGP-STR are unrelated model-specific refactors." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are different models and code paths." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and XLM are different multimodal/text model code paths." + "reason": "ViLT and Flaubert are separate model implementations." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-2 and MGP-STR are unrelated." + "reason": "GPT-2 and MGP-STR are unrelated concrete changes." }, { "left": "pull_request:44017", "right": "pull_request:44161", "accept": false, - "reason": "SegFormer and LongT5 are different model families." + "reason": "Segformer and LongT5 are different model families; only the refactor pattern matches." }, { "left": "pull_request:44017", "right": "pull_request:44073", "accept": false, - "reason": "SegFormer and VisualBert are different implementations; same theme only." + "reason": "Segformer and VisualBERT are unrelated model-specific changes." }, { "left": "pull_request:44017", "right": "pull_request:44722", "accept": false, - "reason": "SegFormer output tracing is unrelated to GPT-J/CodeGen output tracing." + "reason": "Segformer and GPT-J/CodeGen are different code paths." }, { "left": "pull_request:44068", "right": "pull_request:44071", "accept": false, - "reason": "GPT-Neo and MPT are different concrete code paths." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44073", - "accept": false, - "reason": "RegNet and VisualBert are unrelated model implementations." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44018", - "accept": false, - "reason": "SegFormer and GPT-Neo are different model families." - }, - { - "left": "pull_request:44073", - "right": "pull_request:44161", - "accept": false, - "reason": "VisualBert and LongT5 are unrelated code paths." + "reason": "GPT-Neo and MPT are different models." }, { "left": "pull_request:44073", - "right": "pull_request:44101", + "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and XLM are different implementations." + "reason": "VisualBERT and TextNet are different concrete model implementations." }, { "left": "pull_request:44059", "right": "pull_request:44071", "accept": false, - "reason": "GPT-2 and MPT are different models; only the decorator migration theme overlaps." + "reason": "GPT-2 and MPT are separate model code paths." }, { "left": "pull_request:44018", "right": "pull_request:44059", "accept": false, - "reason": "GPT-Neo and GPT-2 are distinct model-specific refactors." + "reason": "GPT-Neo and GPT-2 are different model-specific changes." }, { "left": "pull_request:44056", "right": "pull_request:44101", "accept": false, - "reason": "MPNet and XLM are different code paths." - }, - { - "left": "pull_request:44059", - "right": "pull_request:44066", - "accept": false, - "reason": "GPT-2 versus GPT-J/CodeGen is not the same underlying change." - }, - { - "left": "pull_request:44101", - "right": "pull_request:44161", - "accept": false, - "reason": "XLM and LongT5 are unrelated models." + "reason": "MPNet and Flaubert are unrelated model implementations." }, { "left": "pull_request:43997", "right": "pull_request:44017", "accept": false, - "reason": "RegNet and SegFormer are different image models with different tracing implementations." + "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." + "reason": "RegNet and MPNet are unrelated model-specific refactors." }, { "left": "pull_request:44071", - "right": "pull_request:44129", + "right": "pull_request:44076", "accept": false, - "reason": "MPT and SpeechT5 are different model code paths." + "reason": "MPT and ImageGPT are different concrete changes." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet and XLM are unrelated." + "reason": "TextNet and Flaubert are unrelated model-specific changes." }, { "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "MPNet and LongT5 are different models and changes." + "reason": "MPNet and LongT5 are different models and code paths." }, { "left": "pull_request:43997", "right": "pull_request:44018", "accept": false, - "reason": "RegNet and GPT-Neo are different concrete implementations." + "reason": "RegNet and GPT-Neo are different concrete changes." }, { "left": "pull_request:43997", "right": "pull_request:44161", "accept": false, - "reason": "RegNet and LongT5 are unrelated." + "reason": "RegNet and LongT5 are unrelated model-specific refactors." }, { - "left": "pull_request:44074", + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT and MGP-STR are different code paths." + }, + { + "left": "pull_request:44073", "right": "pull_request:44161", "accept": false, - "reason": "TextNet and LongT5 are unrelated model changes." + "reason": "VisualBERT and LongT5 are unrelated model-specific changes." }, { - "left": "pull_request:44010", - "right": "pull_request:44071", + "left": "pull_request:44073", + "right": "pull_request:44101", "accept": false, - "reason": "SqueezeBert and MPT are different code paths." + "reason": "VisualBERT and Flaubert are different models." }, { - "left": "pull_request:44018", - "right": "pull_request:44129", + "left": "pull_request:44059", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-2 and GPT-J/CodeGen are different model code paths." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", + "accept": false, + "reason": "Flaubert and LongT5 are unrelated model implementations." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", "accept": false, - "reason": "GPT-Neo and SpeechT5 are unrelated implementations." + "reason": "CVT/FNet and Segformer are different concrete changes." }, { "left": "pull_request:44010", - "right": "pull_request:44073", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT and DeBERTa are different model code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are unrelated model-specific refactors." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are different models; same tracing pattern only." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", "accept": false, - "reason": "SqueezeBert and VisualBert are different model-specific refactors." + "reason": "VisualBERT and TextNet are different model-specific changes." }, { "left": "pull_request:44010", "right": "pull_request:44018", "accept": false, - "reason": "SqueezeBert and GPT-Neo are unrelated." + "reason": "SqueezeBERT and GPT-Neo are unrelated concrete code paths." }, { - "left": "pull_request:44044", + "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": "DeBERTa and TextNet are different model families." + "reason": "RegNet and TextNet are unrelated model-specific refactors." }, { - "left": "pull_request:44044", - "right": "pull_request:44129", + "left": "pull_request:44071", + "right": "pull_request:44085", "accept": false, - "reason": "DeBERTa and SpeechT5 are unrelated." + "reason": "MPT and GPT-J are different concrete changes." }, { - "left": "pull_request:44044", - "right": "pull_request:44154", + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are unrelated model-specific changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", "accept": false, - "reason": "DeBERTa and ViTS are different code paths and not the same bug/change." + "reason": "FNet/CVT and LongT5 are different code paths." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and stays conservative about the soft edges. The two accepted duplicate-like pairs (44018/44068 and 44066/44085) are supported by shared model/file overlap and closely aligned output-tracing refactors, and the many rejections appropriately keep different model families separate. The canonical/best-PR reasoning is also consistent with the explicit issue links and the GPT-J/CodeGen overlap around 44722." + "feedback": "The summary is grounded in the packet: the cluster is a set of model-specific output-tracing refactors, #43979 is the shared umbrella issue target, and the chosen canonical PR (#44722) is plausibly the broadest GPT-J/CodeGen representative. The soft-edge accepts are conservative and supported by same-model overlap, with no clear evidence of overstating duplicate status." }, "evaluator_used": true, "retried": false, @@ -25172,12 +25126,12 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "fbfbbefba77c9b2d7fd4ce3ef367f852db486a0dfac3957fe62adf0ca0c92303", + "prepared_review_unit_hash": "4adf6a220c68e8312df275a89dcbb287852c228b0c481cbaf016fd462fa12ed8", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 76, - "serialized_chars": 84119, + "serialized_chars": 84117, "estimated_input_tokens": 21030, "estimated_eval_tokens": 42316 }, @@ -25194,36 +25148,36 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These PRs are all model-specific refactors to adopt the standardized output-capturing/tracing machinery, but they touch different code paths and models, so they are not duplicates of one another.", - "confidence": 0.93, + "summary": "This cluster is a set of model-specific PRs that all migrate different architectures to standardized output-capture patterns, but they are not duplicates of one another. Each touches a different model implementation and would not plausibly be merged into a single fix PR.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44056 is the cleanest representative of the cluster: it explicitly introduces the standardized output-capture pattern in MPNet with the core imports, helper wiring, and _can_record_outputs mapping.", + "canonical_pr_reason": "PR 44129 is the most representative only because it has an explicit issue target and a more complete refactor, but it is still distinct from the other PRs and not a duplicate of them.", "best_issue_reason": null, - "best_pr_reason": "PR 44056 is the best single exemplar of this refactor theme because it is focused, self-contained, and shows the canonical mechanics of the output-tracing migration without extra unrelated surface area.", + "best_pr_reason": "PR 44129 is the strongest standalone representative of the cluster: it has an explicit tracked issue, a concrete code-path change in SpeechT5, and the broadest refactor context among the candidates.", "soft_edge_verdicts": [ { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Both are the same high-level refactor pattern, but they affect different models and different forward paths; that is not the same concrete bug/change and they would not be merged as one PR." + "reason": "Same high-level migration theme, but MPNet and VisualBert are different code paths with model-specific tracing setup. Not the same underlying change and not mergeable as one PR." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "These are separate model migrations to the same framework pattern, but they modify different implementations and test surfaces, so they are not a duplicate change." + "reason": "Both are output-tracing refactors, but they target different models and different implementations. Similar subsystem work is not enough to treat them as duplicates." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are unrelated model code paths; similar output-tracing cleanup does not make them the same underlying change." + "reason": "RegNet and SqueezeBert are unrelated model implementations. The shared decorator migration pattern does not indicate the same bug or change." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 target different architectures and forward signatures, so this is not one mergeable duplicate PR." + "reason": "These are separate model-level refactors in different code paths. The explicit issue link in SpeechT5 does not make it a duplicate of the SqueezeBert PR." } ] }, @@ -25303,274 +25257,274 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "bafbb73e575059e1dc0f62be626d184ab888cbbc4e7612baad4af695a77cd035", + "prepared_review_unit_hash": "309cc696164f51c127e040b5453baa33d7656d9d9b3016954330d6fa1e7da200", "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": 40185, + "estimated_input_tokens": 10047, + "estimated_eval_tokens": 20350 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40177, - "estimated_input_tokens": 10045, - "estimated_eval_tokens": 20346 + "serialized_chars": 40185, + "estimated_input_tokens": 10047, + "estimated_eval_tokens": 20350 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs share a broad theme\u2014refactoring output tracing / `can_return_tuple` handling\u2014but they target different model implementations and code paths, so they are not duplicates.", - "confidence": 0.97, + "summary": "Loose thematic cluster of model-specific output-tracing refactors. The PRs share a similar modernization pattern, but they target different models and code paths, so none look like true duplicates.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "No true duplicate PR stands out; if one must anchor the cluster, PR 44003 is the broadest representative because it spans both `mamba` and `falcon_mamba` and has the largest concrete refactor footprint.", + "canonical_pr_reason": "No true canonical duplicate stands out: each PR is a separate model-specific output-tracing refactor, not the same concrete change. If one must be named, 44025 is the clearest representative because it applies the pattern across two related depth-anything files.", "best_issue_reason": null, - "best_pr_reason": "PR 44003 is the best representative of the cluster theme because it makes the most substantial output-tracing refactor across closely related model code, while the others are narrower per-model changes.", + "best_pr_reason": "PR 44025 is the best representative of the cluster\u2019s theme: it is a substantial, concrete output-tracing refactor and touches two closely related files, making it the most central exemplar among these separate model-specific changes.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Same refactor theme, but `timm_backbone` and `upernet` are separate model code paths; not the same underlying change." + "reason": "Different model implementations (timm_backbone vs upernet); same refactor style, but not the same underlying change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Both are output-tracing refactors, but they touch unrelated model families and cannot be merged as one concrete fix." + "reason": "Different models and forward paths; both are output-tracing cleanups, but not a duplicate bug/change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Different models (`timm_backbone` vs `univnet`) and distinct implementations; similarity is only thematic." + "reason": "Separate model modules (timm_backbone vs univnet); similar cleanup pattern only." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Model-specific refactors in different files; no shared concrete code-path problem." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "`timm_backbone` and `mamba/falcon_mamba` are unrelated code paths, so this is not one duplicate change." + "reason": "timm_backbone and falcon_mamba/mamba are unrelated implementations; same theme, not same change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different architectures and files; not the same fixable code-path issue." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "These are separate model-specific refactors; no shared concrete bug or fix." + "reason": "Separate model-specific tracing updates; no evidence they are the same underlying bug." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "Both refactor output handling, but `mobilenet_v1` and `univnet` are unrelated models." + "reason": "mobilenet_v1 and univnet are different model code paths; similar refactor, not duplicate." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Distinct model implementations; too broad to be the same underlying change." + "reason": "Different models and different forward semantics; not one mergeable duplicate fix." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and files; this is only a shared refactor pattern." + "reason": "Both are output-tracing refactors, but in unrelated model families and separate files." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "`mobilenet_v1` and `vision_encoder_decoder` are separate code paths with no single duplicate fix." + "reason": "mobilenet_v1 vs vision_encoder_decoder; shared modernization pattern only." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Same theme, but `mobilenet_v1` and `speech_encoder_decoder` are unrelated implementations." + "reason": "Different model wrappers and code paths; not the same concrete change." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "`vision_text_dual_encoder` and `univnet` are different model stacks; not the same concrete bug." + "reason": "vision_text_dual_encoder and univnet are unrelated models; same refactor theme, not duplicate." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "These touch different model families and should stay separate." + "reason": "Different model implementations with separate forward logic; not the same bug/change." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; no evidence of a shared code-path defect." + "reason": "vision_text_dual_encoder vs falcon_mamba/mamba are distinct code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different architectures and files; too broad a similarity to count as duplicate." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Similar refactor wording, but `vision_text_dual_encoder` and `depth_anything` are independent changes." + "reason": "Same output-tracing motif, but applied to unrelated model families." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "`univnet` and `upernet` are unrelated model implementations." + "reason": "UnivNet and UperNet are separate models; no shared concrete issue." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Different model families; similar output-tracing cleanup is not enough to merge them." + "reason": "Different model code paths; not plausibly one merged PR." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Separate model-specific refactors with no shared concrete fix." + "reason": "Separate model-specific tracing refactors; not the same underlying change." }, { - "left": "pull_request:44002", - "right": "pull_request:44029", + "left": "pull_request:44001", + "right": "pull_request:44026", "accept": false, - "reason": "`upernet` and `rwkv` are unrelated code paths." + "reason": "UnivNet vs vision_encoder_decoder; similar pattern only." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; not one underlying bug/change." - }, - { - "left": "pull_request:44003", - "right": "pull_request:44028", - "accept": false, - "reason": "`mamba/falcon_mamba` and `superpoint` are separate changes; only the refactor style matches." - }, - { - "left": "pull_request:44003", - "right": "pull_request:44004", - "accept": false, - "reason": "`mamba` and `codegen` are distinct model implementations and cannot reasonably be merged as one fix." - }, - { - "left": "pull_request:43998", - "right": "pull_request:44004", - "accept": false, - "reason": "`timm_backbone` and `codegen` are different code paths; shared wording does not imply duplication." + "reason": "UperNet and falcon_mamba/mamba touch unrelated implementations." }, { - "left": "pull_request:44001", + "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "Different models and concrete behavior; no duplicate change." + "reason": "Different model families and code paths; no duplicate fix here." }, { - "left": "pull_request:44002", - "right": "pull_request:44004", + "left": "pull_request:44003", + "right": "pull_request:44028", "accept": false, - "reason": "`upernet` and `codegen` are unrelated refactors." + "reason": "falcon_mamba/mamba vs superpoint are unrelated model changes." }, { - "left": "pull_request:44000", + "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "These are separate model-specific output tracing changes, not the same fix." + "reason": "CodeGen and Mamba/Falcon Mamba are distinct implementations; not the same concrete change." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Both are wrapper-style refactors, but `depth_anything` and `vision_encoder_decoder` are different code paths." + "reason": "Depth-Anything and vision_encoder_decoder are different model paths; same cleanup style only." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "`depth_anything` and `speech_encoder_decoder` are unrelated models." + "reason": "Different encoder-decoder-style modules, but separate code paths and changes." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Different model implementations; not one underlying bug or change." + "reason": "Depth-anything vs superpoint; not the same underlying bug or refactor target." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "`depth_anything` and `rwkv` target different code paths and should not be merged." + "reason": "Different models (depth-anything vs rwkv); similarity is only at the refactor theme level." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Separate model-specific tracing refactors; no shared concrete defect." + "reason": "Depth-anything and DPR are unrelated model implementations." }, { "left": "pull_request:44026", - "right": "pull_request:44029", + "right": "pull_request:44027", "accept": false, - "reason": "`vision_encoder_decoder` and `rwkv` are unrelated changes." + "reason": "Both are encoder-decoder wrappers, but one is vision and the other speech; distinct concrete code paths." }, { "left": "pull_request:44026", - "right": "pull_request:44027", + "right": "pull_request:44028", "accept": false, - "reason": "Wrapper refactor similarity only; vision and speech encoder-decoder are distinct implementations." + "reason": "vision_encoder_decoder and superpoint are unrelated models; not mergeable as one fix." }, { "left": "pull_request:44026", - "right": "pull_request:44028", + "right": "pull_request:44029", "accept": false, - "reason": "Different model code paths and change sets." + "reason": "Different model families and separate forward wrappers; not duplicates." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "`vision_encoder_decoder` and `dpr` are separate model-specific updates." + "reason": "vision_encoder_decoder vs DPR are separate changes with only a broad theme in common." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "`speech_encoder_decoder` and `superpoint` do not fix the same concrete issue." + "reason": "speech_encoder_decoder and superpoint are unrelated code paths." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Different model families; too broad to merge as duplicates." + "reason": "Different models and different output semantics; not the same fix." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "`speech_encoder_decoder` and `dpr` are unrelated implementations." + "reason": "speech_encoder_decoder and DPR are distinct model implementations." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "`superpoint` and `rwkv` are different code paths with no shared underlying bug." + "reason": "SuperPoint and RWKV are unrelated model families; same style of refactor only." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "These are separate model refactors, not one duplicate PR." + "reason": "SuperPoint and DPR are different modules with no shared concrete bug." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "`rwkv` and `dpr` are unrelated model implementations; similarity is only in refactor style." + "reason": "RWKV and DPR are separate model code paths; not a duplicate change." } ] }, @@ -25600,7 +25554,7 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "cf5ff5353799e44a4da771dae5a8c30889283c7fca50ea7e8867f5f29fb31aa9", + "prepared_review_unit_hash": "c2412163340ff30c8bb3f364e3c18c3a443c6c4fe2a68efcf255afff434f37f4", "component_budget": { "node_count": 2, "item_count": 2, @@ -25622,18 +25576,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated model-addition PRs that only overlap on common Transformers registration/docs scaffolding. They target different models, codepaths, and release tasks, so they should not be deduplicated.", - "confidence": 0.98, + "summary": "These are two separate model-integration PRs: one adds PenguinVL, the other adds Voxtral TTS. They share standard Transformers scaffolding files, but the model codepaths, configs, processors, and tests are distinct, so they should not be treated as duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45401 is the better canonical representative for the cluster because it is tied to a specific tracked issue and has a narrowly scoped, model-specific implementation for Voxtral TTS. PR #44662 is a separate PenguinVL addition and not a duplicate.", + "canonical_pr_reason": "No canonical PR: the PRs implement different models and fix different feature gaps, so there is no single underlying change to merge them under.", "best_issue_reason": null, - "best_pr_reason": "#45401 is the best PR to keep as the representative item since it has an explicit issue target and a clearly bounded deliverable. #44662 is larger but still a distinct PenguinVL model implementation, not the same change.", + "best_pr_reason": "PR 45401 is the better standalone representative because it is tied to an explicit tracking issue and has a narrower, concrete scope around Voxtral TTS support.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different model additions (PenguinVL vs Voxtral TTS) with different filenames, implementations, and goals. Shared auto-registration/docs files are expected for any new model and do not indicate the same underlying change." + "reason": "Reject: PenguinVL and Voxtral TTS are unrelated model additions. The overlap is mostly boilerplate integration files (docs and auto registries), while the model implementations, processors, and tests are for different architectures and use cases." } ] }, @@ -25663,7 +25617,7 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "efbae4c392c0f9b9f253cef713abbc7061bed778622145ed65f7aac88f8eecc5", + "prepared_review_unit_hash": "886ddc9fc4732cfb2e67f525ae3d903a054986dc3a48a4a2e55a9a84ea80e48d", "component_budget": { "node_count": 2, "item_count": 2, @@ -25685,18 +25639,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs are related conceptually around module/kernel fusion, but they target different layers and are not the same change set: #44979 adds a generic Module Fusion API, while #45363 wires in n-to-1 kernel fusion through KernelConfig and model loading integration. They should not be merged as duplicates.", - "confidence": 0.87, + "summary": "Both PRs are about model/module fusion, but they target different layers: #44979 introduces a generic Module Fusion API, while #45363 adds KernelConfig-based n-to-1 kernel fusion and integration plumbing. They are related, but not the same underlying change and not mergeable as one duplicate PR.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #44979 is the better canonical representative because it introduces the core module-fusion abstraction and API; the other PR is a later, more specific integration on top of that idea.", + "canonical_pr_reason": "#45363 is the more concrete, user-facing fusion feature with broader integration changes; #44979 is a separate generic API layer rather than the same fix/change.", "best_issue_reason": null, - "best_pr_reason": "PR #44979 is the best overall representative of the cluster since it is the foundational API change, whereas #45363 is a narrower application of fusion to kernel loading/configuration.", + "best_pr_reason": "#45363 is the strongest representative of the cluster theme because it implements an end-to-end fusion path across multiple files, whereas #44979 is a distinct abstraction/API proposal.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Not the same underlying change: #44979 adds a standalone Module Fusion API, while #45363 implements n-to-1 kernel fusion via KernelConfig and related model/integration code. They could coexist and are not plausibly merged into one PR as duplicates." + "reason": "Same broad fusion area, but different abstractions and code paths: generic module fusion API vs KernelConfig-based kernel fusion. Not the same concrete change and not a single plausible merged PR." } ] }, @@ -25726,12 +25680,12 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "77e84236883c6b340110131660e35181bba781e1b035cd1b5b927fda9bdbd368", + "prepared_review_unit_hash": "67270ed2190a8f0b83f02d0adc622c2ee4ad6fddec79692b54ba8c1132b0c361", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4413, + "serialized_chars": 4414, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -25739,7 +25693,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4413, + "serialized_chars": 4414, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -25748,18 +25702,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests address the same AttributeError in `_patch_mistral_regex` by switching Mistral pre-tokenizer access from `backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; the second PR adds a test but does not change the underlying fix.", - "confidence": 0.97, - "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45317 is the better canonical PR because it fixes the same bug as 45086 and also includes a test, making it the more complete representative of the change.", - "best_issue_reason": null, - "best_pr_reason": "PR 45317 is the best overall PR for the cluster because it covers the same concrete code-path fix and adds validation, while remaining the same narrowly scoped change.", + "summary": "Both PRs address the same AttributeError in `_patch_mistral_regex` by changing the pre-tokenizer access path; PR 45317 is the more complete variant because it adds tests.", + "confidence": 0.98, + "canonical_issue_reason": "Both PRs explicitly target issue #45081, which appears to be the shared bug report for this exact failure.", + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same code fix as 45086 plus test coverage, making it the more complete implementation.", + "best_issue_reason": "No issue item is present in the cluster, but issue #45081 is the common upstream bug being fixed by both PRs.", + "best_pr_reason": "PR 45317 is the best PR candidate since it fixes the same concrete code path as 45086 and adds validation tests.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying AttributeError in `_patch_mistral_regex` on the same file and code path; the later PR is the same fix with added test coverage, so they are duplicates for triage." + "reason": "Same underlying AttributeError fix in the same function and same file; 45317 is essentially the same change with added tests, so these are duplicates/same change family." } ] }, @@ -25789,12 +25743,12 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "06c8e5a51a938f78d6986701c0fca019612b6ed05f4d31778e7baf7aece6a605", + "prepared_review_unit_hash": "35aeec7833ce9e3634261a4569d530605fe82c6a498007d7a4ab9d0f88a7a5e8", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6066, + "serialized_chars": 6065, "estimated_input_tokens": 1517, "estimated_eval_tokens": 3290 }, @@ -25802,7 +25756,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6066, + "serialized_chars": 6065, "estimated_input_tokens": 1517, "estimated_eval_tokens": 3290 }, @@ -25811,18 +25765,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs touch NemotronH, but they are not the same change: #45115 is a focused conversion/refactor mapping update, while #45569 is the broader implementation split introducing NemotronH Dense/Sparse models, docs, auto-mappings, and tests. They are related but not duplicate-worthy as a single underlying fix/change.", - "confidence": 0.87, + "summary": "The two PRs are related to NemotronH but they are not duplicates: one is a narrow conversion-mapping / inheritance refactor, while the other is a broader split into NemotronHDense and NemotronHSparse with docs and auto-mapping updates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "#45569 is the more complete and central PR: it introduces the actual model split and associated library wiring, docs, and tests, whereas #45115 is a narrower helper/refactor around conversion mappings.", + "canonical_pr_reason": "If a representative PR must be chosen, 45569 is the broader, more central architectural change; 45115 is a narrower follow-up/refactor focused on checkpoint conversion mappings.", "best_issue_reason": null, - "best_pr_reason": "#45569 best represents the main functional change to NemotronH and its variants; #45115 is auxiliary infrastructure for checkpoint conversion rather than the core implementation.", + "best_pr_reason": "45569 is the better representative PR for the cluster because it covers the substantive model-family split and public-facing integration work, whereas 45115 is limited to rename/mapping adjustments.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Related NemotronH work, but not the same concrete bug/change. #45115 only adds conversion-mapping/refactor support, while #45569 implements the broader model split and API/docs/test updates; they would not naturally be merged as one duplicate PR." + "reason": "Different underlying changes: 45115 updates NemotronH conversion mapping and compatibility renames; 45569 introduces separate dense/sparse NemotronH variants plus docs, auto-mapping, and tests. Shared files are incidental to the same model family, not evidence of the same fix." } ] }, @@ -25852,7 +25806,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "22ed84b3e447b2cd0aea8002e18a4dc49f87577750adc8f8ced68a51855fca17", + "prepared_review_unit_hash": "8ac2f336b92588b14df661a0ac7abeef715a3cc9b9295ea8c9e26d16fa0a356d", "component_budget": { "node_count": 2, "item_count": 2, @@ -25874,18 +25828,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about heterogeneous configuration support, but they are not the same change: #45333 only adds per-layer config support, while #45332 is a larger end-to-end heterogeneous model support PR that also touches modeling/cache/masking/integration paths.", - "confidence": 0.92, + "summary": "Both PRs are about heterogeneous model support, but they are not the same change: #45333 is a focused per-layer configuration PR, while #45332 is a much broader end-to-end heterogeneous modeling feature that also touches cache, masking, integrations, and model wiring. They look related and likely sequential, not duplicates.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "#45332 is the broader canonical PR because it implements the full heterogeneous model support stack, not just config plumbing, and therefore subsumes the narrower config-only work in #45333.", + "canonical_pr_reason": "#45332 is the broader umbrella change and best represents the cluster because it includes the configuration work plus the actual heterogeneous modeling/integration implementation.", "best_issue_reason": null, - "best_pr_reason": "#45332 is the best representative of the cluster since it covers the full feature area and is the more comprehensive implementation; #45333 is a narrower subset focused on configuration only.", + "best_pr_reason": "#45332 is the strongest standalone PR for this cluster: it covers the full feature surface and is more likely to be the principal implementation users care about.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related but not the same concrete change. #45333 is config-only, while #45332 adds broader modeling/runtime support across multiple code paths, so they should not be merged as duplicates." + "reason": "Not the same underlying PR: #45333 is config-only, while #45332 adds a larger heterogeneous modeling stack across cache/masking/integrations. They share some files and theme, but they are not mergeable as one duplicate change." } ] }, @@ -25915,7 +25869,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "f1343316dfdfe1d25dcc240089a5fbdac01f50b898efadc4a80e78e78661dc5c", + "prepared_review_unit_hash": "cddcef27f5dcbaa652cb0a3b9cb445425898798001ec5cf8578e8b3af4768f56", "component_budget": { "node_count": 2, "item_count": 2, @@ -25937,18 +25891,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related typing-focused PRs, but they are not the same change: one is a broad type-checking sweep across many Transformers files, while the other adds specific typing protocols and helpers for modeling_utils/PEFT integration. They should not be merged as duplicates.", - "confidence": 0.88, + "summary": "The two PRs are related by typing work and a few shared files, but they address different scopes: one is a broad type-checking sweep across many `src/transformers/*` files, while the other adds targeted typing helpers/protocols and narrower model-utils integration. They do not look like the same underlying change.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:45425 is the more focused, self-contained typing change; pull_request:45415 is a much broader project-wide type-checking sweep.", + "canonical_pr_reason": "PR 45425 is the more coherent representative of the cluster: it introduces focused typing abstractions and concrete typing support around `modeling_utils`/related helpers, rather than a wide mechanical type-checking pass.", "best_issue_reason": null, - "best_pr_reason": "pull_request:45425 is the better representative PR because it targets a concrete typing compatibility gap with narrower scope and clearer behavioral intent.", + "best_pr_reason": "PR 45425 is the best single representative because it is narrower and more semantically focused, making it easier to understand as a distinct change than the broad repo-wide typing cleanup in PR 45415.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Both are typing-related and share some helper files, but they address different scopes: 45415 is a broad type-checking enablement pass, while 45425 adds specific protocols/type aliases for modeling_utils and PEFT-related code. They do not look like the same underlying fix and would not naturally merge into one PR." + "reason": "Both are typing-related and share a few files, but they solve different problems: 45415 is a broad type-checking enablement pass across many modules, while 45425 adds specific typing protocols/aliases and targeted fixes. They are not the same concrete bug/change and would not naturally merge into one PR." } ] }, @@ -25978,7 +25932,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "17a15f7597f031f4515dbda10e8cff44aeea1b50fac54247dd6671c15810f266", + "prepared_review_unit_hash": "9c51e3f56fb496d68dfa9de2179c5c31038f551b081dcf012d5295582991c893", "component_budget": { "node_count": 2, "item_count": 2, @@ -26000,18 +25954,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related EP-loading fixes but they target different failure modes and code paths: one addresses DeepSpeed ZeRO-3 loading via accelerate launch, the other fixes FSDP2/rank-0 broadcast overwriting experts. They should not be deduplicated.", - "confidence": 0.92, + "summary": "The two PRs are related by expert-parallel loading, but they fix different bugs: one is about preserving/reading `distributed_config` during Accelerate/ZeRO-3 loading, while the other fixes FSDP2 expert parameters being overwritten after sharding. They are adjacent, not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45662 is the better canonical PR because it targets the more specific and higher-severity defect (experts being silently overwritten during distributed loading) and includes the broader set of coordinated code changes needed for that path.", + "canonical_pr_reason": "PR #45662 is the better canonical representative for the cluster because it addresses a concrete, self-contained loading bug in the parameter assignment path; PR #45548 is a separate configuration/serialization fix and does not collapse into the same change.", "best_issue_reason": null, - "best_pr_reason": "PR #45662 is the strongest representative of the cluster because it fixes a concrete correctness bug in distributed expert loading, with changes spanning the relevant loading and MoE/tensor-parallel paths. PR #45548 is narrower and addresses a different distributed-loading scenario.", + "best_pr_reason": "PR #45662 is the strongest standalone fix: it targets a specific code path (`set_param_for_module` / post-shard wrapping) and has a clear, reproducible failure mode. PR #45548 is related but covers a different EP/DeepSpeed configuration issue.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Both mention expert parallelism and touch modeling_utils.py, but they fix different bugs: #45548 is about EP + DeepSpeed ZeRO-3 loading with accelerate launch, while #45662 is about EP + FSDP2 rank-0 broadcast overwriting experts. The code paths and failure modes are distinct, so they are not duplicates." + "reason": "Both are EP-loading related, but they fix different problems in different paths: #45548 preserves `distributed_config` and adds `has_ep`, while #45662 prevents expert tensors from being overwritten after sharding via `post_shard_wrap`. Not the same underlying bug." } ] }, @@ -26041,40 +25995,40 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "53bc560079635b87d2809179945b5420d961331c3470ef77a86f720f2b5d3a47", + "prepared_review_unit_hash": "783efdaa4c40f41bfd54ab812a92843dcf7ba2065e62e98bd9608e67984297a3", "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 target the same xdist/captured_info debugging-log problem in `testing_utils`, and 45645 appears to be the more complete version by also fixing CI log collection.", - "confidence": 0.89, + "summary": "Both pull requests target the same xdist-safe captured_info logging problem under issue #45561; 45645 is a broader superset of 45639 that also updates CI log collection.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the better canonical representative because it includes the core xdist-safe `captured_info` fix plus the workflow change needed to preserve and display all generated debug logs in CI.", + "canonical_pr_reason": "PR #45645 is the better canonical representative because it includes the core testing_utils xdist-safe fix plus the CI workflow change needed to preserve and collect the split log files.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 is the strongest single choice: it addresses the underlying collision bug and the downstream CI artifact handling together, making it the most complete fix in the cluster.", + "best_pr_reason": "PR #45645 is the strongest overall match for the cluster: it addresses the same bug as #45639 and adds the downstream artifact-handling changes needed to make the fix complete in CI.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Both PRs fix the same concrete `captured_info`/xdist collision path in `testing_utils`, share the same tracked issue, and could plausibly be merged into one combined PR." + "reason": "They fix the same underlying captured_info/xdist collision issue in the same code path, and #45645 plausibly subsumes #45639 with additional workflow support." } ] }, @@ -26170,7 +26124,7 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "5acdac20e1630fa2fd7e6b5cbfaebb49bd56534a1ae70402dcbf000c75e296cf", + "prepared_review_unit_hash": "2ca6a640f3586da2b050586c5342f5ec0325257d090037859fb0a09d5c2d73fb", "component_budget": { "node_count": 2, "item_count": 2, @@ -26192,18 +26146,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two open PRs in the distributed stack, but they are not the same change: #46126 is a focused EP/Trainer integration feature, while #46269 is a broad distributed branch base touching many unrelated subsystems. The overlap in a few filenames is too coarse to treat them as duplicates.", - "confidence": 0.93, + "summary": "The two pull requests overlap on distributed code, but they are not the same change: one is a focused EP/trainer integration feature, while the other is a broad distributed branch base spanning many unrelated files. The soft duplicate edge should be rejected.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "#46269 is the broader umbrella branch that includes the overlapping distributed files and looks like the main integration base for that line of work.", + "canonical_pr_reason": "PR #46126 is the more specific, self-contained change set: it targets EP routing behavior and trainer integration on top of DistributedConfig, with a narrow set of touched files and a clear feature intent. PR #46269 is much broader and looks like a base/rollup branch rather than a single concrete fix.", "best_issue_reason": null, - "best_pr_reason": "#46126 is the better representative PR for a concrete code change because it has a single, well-scoped feature intent and clearer mergeable behavior.", + "best_pr_reason": "PR #46126 is the best representative PR for the cluster because it has the clearest single-purpose change and the tightest scope. PR #46269 is too wide-ranging to serve as the canonical example for this cluster.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "They share distributed filenames, but the PRs are not the same underlying fix/change: #46126 is EP + Trainer integration, while #46269 is a broad distributed branch base spanning FSDP, MoE, tensor parallel, modeling, and CI updates." + "reason": "Reject: they share distributed filenames, but the change intent is different. #46126 is an EP + trainer integration feature with router-parallel logic; #46269 is a broad distributed branch base touching many subsystems, so they do not look like the same underlying PR." } ] }, @@ -26271,220 +26225,220 @@ "pull_request:46504|pull_request:46716", "pull_request:46432|pull_request:46716" ], - "prepared_review_unit_hash": "5c91bc606bc088ca01913869d3f45b3b591b23b34c4a6fd4279278c96d97845d", + "prepared_review_unit_hash": "0d9e6be9906837afab8451f5bac450132ba2805572d2cd101d36b79e6d816443", "component_budget": { "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26678, - "estimated_input_tokens": 6670, - "estimated_eval_tokens": 13596 + "serialized_chars": 26665, + "estimated_input_tokens": 6667, + "estimated_eval_tokens": 13590 }, "budget": { "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26678, - "estimated_input_tokens": 6670, - "estimated_eval_tokens": 13596 + "serialized_chars": 26665, + "estimated_input_tokens": 6667, + "estimated_eval_tokens": 13590 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a set of small, mostly independent workflow-maintenance PRs. The similarity is largely superficial (same chore-style titles, nearby YAML edits, or same broad CI/security area), but the diffs change different files and different behaviors, so they are not true duplicates.", - "confidence": 0.91, + "summary": "These PRs are a broad set of workflow-maintenance edits, not true duplicates. The only tight cluster is several changes to pr-ci-security-gate.yml, but they modify different parts of that workflow and should not be merged as one duplicate change.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR stands out: the items span unrelated workflow files, and even the multiple pr-ci-security-gate edits address different concerns rather than the same concrete change.", + "canonical_pr_reason": "No clear canonical PR: the set splits into unrelated workflow tweaks, and even the pr-ci-security-gate edits are distinct changes to the same file rather than one duplicate fix.", "best_issue_reason": null, - "best_pr_reason": "PR 46554 is the closest representative of the only tight subgroup because it makes the most substantive security-gate workflow change and shares the exact file with several other candidates, but it still should not be treated as a duplicate target for the others.", + "best_pr_reason": "PR 46554 is the most substantive and self-contained change in the security-gate subgroup, but it is still not a duplicate of the later permissions/reusable-workflow/secrets-only updates.", "soft_edge_verdicts": [ { "left": "pull_request:46138", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and unrelated changes: self-comment CI stdin handling vs PR CI security-gate logic." + "reason": "Different workflow files and different behavior; both are CI chores, but they do not fix the same underlying change." }, { "left": "pull_request:46138", "right": "pull_request:46558", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Same broad CI area, but one changes self-comment message piping while the other adjusts PR security-gate permissions and approval flow." }, { "left": "pull_request:46138", "right": "pull_request:46563", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Unrelated workflow changes; shared CI-maintenance style is not enough for duplication." }, { "left": "pull_request:46138", "right": "pull_request:46716", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflow and different concern; one updates comment handling, the other tightens security-gate secrets." }, { "left": "pull_request:46138", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Both are shell-safety tweaks, but they affect different workflows and different code paths." }, { "left": "pull_request:46432", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and unrelated changes; ssh-runner hardening is not the same change as benchmark dispatch support." + "reason": "Different workflows and unrelated runtime behavior; no shared underlying bug or fix." }, { "left": "pull_request:46432", "right": "pull_request:46501", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "One updates ssh-runner setup, the other changes anti-slop permissions; not the same change." }, { "left": "pull_request:46432", "right": "pull_request:46502", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflow files and unrelated logic." }, { "left": "pull_request:46432", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "SSH runner hardening is unrelated to pr-ci security-gate logic." }, { "left": "pull_request:46432", "right": "pull_request:46558", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflow concerns; one manages SSH runner env, the other modifies PR CI security-gate behavior." }, { "left": "pull_request:46501", "right": "pull_request:46503", "accept": false, - "reason": "Both are workflow chores, but they touch different CI jobs and fix different behavior." + "reason": "Both touch workflows, but they change different jobs and unrelated behavior." }, { "left": "pull_request:46501", "right": "pull_request:46502", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflow files with unrelated fixes." }, { "left": "pull_request:46501", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "No same underlying bug/change; permissions tweak vs shell input handling in a separate workflow." }, { "left": "pull_request:46501", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflows and different fixes." }, { "left": "pull_request:46502", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Both are workflow chores, but they address different workflows and behaviors." }, { "left": "pull_request:46502", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different files and different logic; not a duplicate fix." }, { "left": "pull_request:46502", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Unrelated workflow changes." }, { "left": "pull_request:46503", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflow files and different behavior; no common underlying change." }, { "left": "pull_request:46503", "right": "pull_request:46554", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Benchmark dispatch support is unrelated to PR security-gate changes." }, { "left": "pull_request:46504", "right": "pull_request:46554", "accept": false, - "reason": "Same broad CI area, but different files and different fixes; not the same underlying bug/change." + "reason": "Similar shell-hardening theme, but they affect different workflows and code paths." }, { "left": "pull_request:46504", "right": "pull_request:46558", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Different workflows and distinct fixes." }, { "left": "pull_request:46554", "right": "pull_request:46558", "accept": false, - "reason": "Same file, but the diffs fix different security-gate concerns and could not plausibly be merged as one duplicate PR." + "reason": "Same file, but different underlying changes: one adds security checks, the other changes workflow permissions/approval gating." }, { "left": "pull_request:46554", "right": "pull_request:46563", "accept": false, - "reason": "Same file, but one adds checkout/checks while the other adds safety comments and workflow wiring; not the same change." + "reason": "Both touch pr-ci-security-gate.yml, but they modify different execution models and are not the same fix." }, { "left": "pull_request:46554", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but one adds new gate logic while the other only changes secrets handling for the reusable workflow." + "reason": "Same workflow file, but one adds code-level checks while the other only changes secrets handling." }, { "left": "pull_request:46558", "right": "pull_request:46563", "accept": false, - "reason": "Same file, but one changes permissions/approval flow and the other adds explanatory comments; not the same underlying fix." + "reason": "Different edits to the same workflow file; not the same underlying bug/change." }, { "left": "pull_request:46558", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but different security-gate edits with different intents." + "reason": "Distinct security-gate tweaks: permissions/approval logic vs secrets inheritance." }, { "left": "pull_request:46563", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but one is documentation/safety-note text and the other changes secrets propagation." + "reason": "Different changes in the same workflow file, but not one duplicate fix." }, { "left": "pull_request:46504", "right": "pull_request:46563", "accept": false, - "reason": "Different files and unrelated workflow behavior." + "reason": "Different workflows and unrelated behavior." }, { "left": "pull_request:46432", "right": "pull_request:46563", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "SSH runner hardening and PR security-gate documentation are unrelated." }, { "left": "pull_request:46504", "right": "pull_request:46716", "accept": false, - "reason": "Different files and unrelated workflow behavior." + "reason": "Different workflow files and different concerns." }, { "left": "pull_request:46432", "right": "pull_request:46716", "accept": false, - "reason": "Different workflow files and unrelated changes." + "reason": "Unrelated workflow changes; no shared code-path problem." } ] }, @@ -26577,12 +26531,12 @@ "soft_pairs": [ "pull_request:46310|pull_request:47162" ], - "prepared_review_unit_hash": "e7fcab596167d6274355a298507ef8f31f15699c523fc6bc7eafb76872399ff6", + "prepared_review_unit_hash": "b6c0f043a84c4fe76d1bab1c61daf066cbfad2d7e8c98554373caccc0cb4bfcf", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4881, + "serialized_chars": 4883, "estimated_input_tokens": 1221, "estimated_eval_tokens": 2698 }, @@ -26590,7 +26544,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4881, + "serialized_chars": 4883, "estimated_input_tokens": 1221, "estimated_eval_tokens": 2698 }, @@ -26599,18 +26553,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs fix the same off-by-one in the MoonshineStreaming sliding-window right mask; PR 47162 is the broader, newer superset that also adds the Gemma4 fix and tests.", - "confidence": 0.88, + "summary": "Both PRs fix the same off-by-one error in the MoonshineStreaming sliding window right bound; PR 47162 is the more complete follow-up because it includes the Moonshine fix plus tests and an additional Gemma4 instance of the same bug.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47162 is the better canonical PR because it contains the same MoonshineStreaming fix as 46310 and additionally covers the same bug pattern in Gemma4 plus tests, making it the more complete change set.", + "canonical_pr_reason": "PR 47162 is the best canonical PR because it contains the same MoonshineStreaming fix as PR 46310, adds tests, and appears to be the more complete version of the patch set.", "best_issue_reason": null, - "best_pr_reason": "PR 47162 is the strongest representative of the bugfix because it addresses the concrete off-by-one window-boundary issue in MoonshineStreaming and includes the related companion fix in Gemma4.", + "best_pr_reason": "PR 47162 is the strongest representative of the cluster: it fixes the same concrete off-by-one bug in MoonshineStreaming and also validates the behavior with tests.", "soft_edge_verdicts": [ { "left": "pull_request:46310", "right": "pull_request:47162", "accept": true, - "reason": "Same underlying MoonshineStreaming right-window off-by-one fix with identical code hunk; 47162 is a superset rather than a different change." + "reason": "Both PRs correct the same right-window off-by-one in sliding_window_mask_function for MoonshineStreaming; 47162 is a superset rather than a different change, so they are close enough to treat as the same underlying fix." } ] }, @@ -26640,7 +26594,7 @@ "soft_pairs": [ "pull_request:46574|pull_request:46580" ], - "prepared_review_unit_hash": "6d03fbd890bc65ae5d8e81f29489d335685ef9f4d60f931860ea65fbb24997c0", + "prepared_review_unit_hash": "20719eedad56e6d08f87c43bb1395627a2c86dce7f938b1f206eef095ce5bde7", "component_budget": { "node_count": 2, "item_count": 2, @@ -26662,18 +26616,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs in different areas: one fixes CodeLlama decode stripping of real leading spaces, the other fixes DeepSeek-Coder/DeepSeek-LLM tokenizer-class misloading in AutoTokenizer. They do not appear to be the same underlying change or bug.", + "summary": "These are both tokenizer-related PRs, but they address different problems in different code paths: CodeLlama decode stripping real leading spaces vs. DeepSeek checkpoint-to-tokenizer-class selection in auto-loading.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single canonical PR stands out because the two PRs fix unrelated tokenizer bugs in separate modules and could not plausibly be merged into one change.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "Neither PR is a good global representative for the cluster since there is no shared underlying bug or change; each is only representative of its own distinct fix.", "soft_edge_verdicts": [ { "left": "pull_request:46574", "right": "pull_request:46580", "accept": false, - "reason": "Different failure modes and code paths: CodeLlama decoder setup in its tokenizer implementation vs AutoTokenizer checkpoint-prefix routing for DeepSeek checkpoints. They are not mergeable as one PR and do not fix the same underlying bug." + "reason": "Different failure modes and code paths: 46574 changes CodeLlama decode behavior in a model-specific tokenizer file, while 46580 changes auto tokenizer class resolution for specific DeepSeek checkpoints. They do not look like the same underlying bug or a mergeable single PR." } ] }, @@ -26706,7 +26660,7 @@ "pull_request:47134|pull_request:47215", "pull_request:46696|pull_request:47134" ], - "prepared_review_unit_hash": "23aa8254d1a2652d4008825462b4c528c7a5fc5424ce6c76de7be79706feb415", + "prepared_review_unit_hash": "bb884b30e85cd2ceef95eede3c27b196fab906fe003e0b230d86edbd296caeea", "component_budget": { "node_count": 3, "item_count": 3, @@ -26728,39 +26682,39 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "47215 is the best canonical PR for the flash-attn performance regression family: it is the most complete fix on the unpadding/max-seqlen path and adds broader coverage. 47134 is an earlier version of that same fix, while 46696 is a separate helper-path change in the same file and issue family; it is related enough to 47215 to keep in the soft-duplicate cluster, but not close enough to be the canonical representative.", - "confidence": 0.78, + "summary": "These PRs are all about flash-attention long-context performance regressions, but they split into two concrete code paths: the unpad/max_seqlen scalar handling and the kwargs-processing/tracing path. PR 47215 is the most complete and likely final fix; 47134 is a partial earlier version of the same unpad regression. 46696 is related but addresses a different helper path, so it is not a duplicate of 47134.", + "confidence": 0.69, "canonical_issue_reason": null, - "canonical_pr_reason": "47215 is the strongest representative: it applies the compile-aware .item() handling to the unpadding/max-seqlen regression path, includes broader validation, and reads like the most complete version of the fix family.", + "canonical_pr_reason": "PR 47215 is the best canonical PR: it is the latest and most complete fix, uses the compile-aware guard, and includes the broader test/benchmark follow-through for the same flash-attention regression.", "best_issue_reason": null, - "best_pr_reason": "47215 best captures the final intended fix for the FA2 long-context performance regression, while 47134 is an earlier narrower variant and 46696 addresses a different helper path.", + "best_pr_reason": "PR 47215 is the strongest representative because it most fully addresses the regression and appears to subsume the earlier partial fix in 47134.", "soft_edge_verdicts": [ { "left": "pull_request:46696", "right": "pull_request:47215", "accept": true, - "reason": "They target the same issue and same file, and both address scalar-sync/performance behavior in flash-attn utilities. The concrete helper differs, but they are close enough to be in the same bugfix family." + "reason": "Both target the same flash-attention long-context regression in the same file and issue thread; 47215 looks like the broader/finalized version of the same fix family." }, { "left": "pull_request:47134", "right": "pull_request:47215", "accept": true, - "reason": "Both fix the unpadding/max-seqlen performance regression in the same code path; 47215 is a refined successor with compile-aware guarding and added coverage." + "reason": "These patch the same unpad/max_seqlen performance bug in the same helper functions; 47215 is a refined, compile-aware replacement for the earlier partial fix." }, { "left": "pull_request:46696", "right": "pull_request:47134", "accept": false, - "reason": "They share the file and issue target, but the fixes are on different helper paths (_process_flash_attention_kwargs vs. unpad/max-seqlen). That is related, but not the same concrete change." + "reason": "They touch different concrete code paths in flash-attention utils: one fixes kwargs/tracing scalar handling, the other fixes unpad/max_seqlen materialization. Related, but not the same bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The summary is mostly grounded, but the soft-edge verdict for pull_request:46696 vs pull_request:47215 is too strong. The packet shows both PRs target the same issue (46693) and share the same file, and the pair is marked with soft_similarity/shared_issue_target evidence. Saying they 'would not be merged as one PR' overstates the separation; they look related enough to stay in the same bugfix family, even if they touch different functions. The 47134 vs 47215 relationship is plausible, but the 46696 vs 47215 rejection is not conservative enough." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. It correctly distinguishes the two concrete flash-attention paths: kwargs/tracing scalar handling in 46696 versus unpad/max_seqlen materialization in 47134/47215. The canonical-choice rationale for 47215 is supported by the later timestamp, broader file coverage, and compile-aware guard. The accepted soft pairs are also plausible mergeable same-bugfix candidates, and the rejected 46696\u219447134 pair is appropriately not overstated as a duplicate." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -26784,40 +26738,40 @@ "soft_pairs": [ "pull_request:46773|pull_request:47087" ], - "prepared_review_unit_hash": "ba85c1853d04335d3490b9f263cf6a5b895b3fe7a9bd4ea10b176fac38860f76", + "prepared_review_unit_hash": "aee84dd9f4239c8957485bfa12db9c287541b107804d186e6b2993c105c18739", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5432, - "estimated_input_tokens": 1358, - "estimated_eval_tokens": 2972 + "serialized_chars": 5433, + "estimated_input_tokens": 1359, + "estimated_eval_tokens": 2974 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5432, - "estimated_input_tokens": 1358, - "estimated_eval_tokens": 2972 + "serialized_chars": 5433, + "estimated_input_tokens": 1359, + "estimated_eval_tokens": 2974 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related by masking bugs, but they fix different code paths: one is linear-attention padding-mask handling for batch-size-1, the other is recurrent-mask handling during continued/cached forwards. They should not be merged as duplicates.", + "summary": "Two open PRs touch padding-mask handling in attention-like models, but they fix different bugs: one restores masking for batch size 1 in linear-attention models, while the other fixes recurrent masking during continued forwards/chunked prefill/cache continuation. They are related in theme but not the same underlying change.", "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47087 is the better cluster representative because it centers the shared masking utility, has an explicit issue target, and covers the recurrent/cached-forward padding-mask bug across multiple recurrent-model consumers.", + "canonical_pr_reason": "PR 47087 is the better canonical PR because it has a clear, specific bug statement, an explicit issue target, and a focused fix in masking_utils with tests covering recurrent continued-forward behavior.", "best_issue_reason": null, - "best_pr_reason": "PR 47087 is the strongest standalone PR in this cluster: it addresses a concrete recurrent-layer masking regression with clear scope, tests, and an explicit issue link.", + "best_pr_reason": "PR 46773 is also a plausible representative for the linear-attention padding-mask bug, but it is narrower and less directly tied to the recurrent-mask behavior than 47087. For this cluster, 47087 is the strongest overall representative of the masking regression theme.", "soft_edge_verdicts": [ { "left": "pull_request:46773", "right": "pull_request:47087", "accept": false, - "reason": "Different underlying bugs and code paths: 46773 fixes linear-attention padding-mask application when batch size is 1, while 47087 fixes recurrent masking being skipped on continued/cached forwards. They are not mergeable as one PR." + "reason": "Both concern padding masks, but 46773 fixes a batch-size-1 masking gate in linear-attention model forward paths, while 47087 fixes mask trimming/skipping for continued forwards in recurrent attention. Different code paths and failure modes; not mergeable as one PR." } ] }, @@ -26854,64 +26808,64 @@ "pull_request:47074|pull_request:47195", "pull_request:47195|pull_request:47197" ], - "prepared_review_unit_hash": "7fefafa568f7461cdfcacbdff25eafa4f05d49a394f71e5c1b97f416db965c1c", + "prepared_review_unit_hash": "5c41ec08009a21b305c11bdf4b96e53c09f8f785828ee07d7f373fb66ac77de4", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 11472, - "estimated_input_tokens": 2868, - "estimated_eval_tokens": 5992 + "serialized_chars": 11473, + "estimated_input_tokens": 2869, + "estimated_eval_tokens": 5994 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 11472, - "estimated_input_tokens": 2868, - "estimated_eval_tokens": 5992 + "serialized_chars": 11473, + "estimated_input_tokens": 2869, + "estimated_eval_tokens": 5994 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are all about different bugs in different code paths (multimodal input routing, placeholder validation, stop-sequence handling, batched postprocessing, and generation length capping). They are related only at a very broad subsystem level, so none should be deduplicated or merged together.", - "confidence": 0.96, + "summary": "These are all transformer multimodal/generation fixes, but they address different concrete bugs in different code paths; none of the soft pairs look like true duplicates.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "No PR here is a duplicate of another. If one must be chosen as the representative, PR #47074 is the broadest concrete multimodal pipeline fix, but it does not subsume the others.", + "canonical_pr_reason": "PR 47074 is the best representative if one must be chosen: it fixes a concrete, user-facing multimodal pipeline input-normalization bug and is the broadest of the set.", "best_issue_reason": null, - "best_pr_reason": "PR #47074 is the best single representative because it fixes a clear, user-facing multimodal input-key bug in a core pipeline and has focused, directly relevant tests; the other PRs are narrower and solve unrelated issues.", + "best_pr_reason": "PR 47074 is the strongest cluster representative because it targets the most general pipeline-level behavior and has the clearest, concrete bug fix among the group.", "soft_edge_verdicts": [ { "left": "pull_request:47032", "right": "pull_request:47197", "accept": false, - "reason": "Different failures: 47032 changes stop-sequence handling in image-text-to-text generation, while 47197 fixes batched decoding in Qwen Omni processors." + "reason": "Different bugs in different components: stop_sequence handling in image-text-to-text vs batched decode/postprocessing in Qwen Omni." }, { "left": "pull_request:47074", "right": "pull_request:47076", "accept": false, - "reason": "Both are multimodal, but one normalizes AnyToAnyPipeline input keys and the other validates Gemma4 audio placeholder batching; they are separate bugs and code paths." + "reason": "Both are multimodal, but one normalizes AnyToAnyPipeline input keys while the other validates Gemma4 audio placeholders; distinct code paths and failure modes." }, { "left": "pull_request:47074", - "right": "pull_request:47197", + "right": "pull_request:47195", "accept": false, - "reason": "47074 fixes pipeline input-key mapping, whereas 47197 fixes postprocessing of generated outputs; they are not the same underlying defect." + "reason": "Unrelated fixes: multimodal input-key normalization vs generation candidate length-capping in the lookup decoder." }, { "left": "pull_request:47074", - "right": "pull_request:47195", + "right": "pull_request:47197", "accept": false, - "reason": "Unrelated areas: multimodal pipeline input normalization versus generation candidate length-capping logic." + "reason": "Separate issues in separate subsystems: AnyToAnyPipeline input mapping vs Qwen Omni text postprocessing." }, { "left": "pull_request:47195", "right": "pull_request:47197", "accept": false, - "reason": "Different subsystems and symptoms: prompt lookup length budgeting in generation versus batched text postprocessing in Qwen Omni." + "reason": "Generation candidate selection length cap and Qwen Omni batched decoding are not the same underlying bug." } ] }, @@ -26941,12 +26895,12 @@ "soft_pairs": [ "pull_request:47105|pull_request:47129" ], - "prepared_review_unit_hash": "9b225eedde7592424668337e39dc4b6ab2c0079311ae73cab4546baf751640d0", + "prepared_review_unit_hash": "7e2cd2353573798238b25945542e01d22aa7df063e18b0fe22acae597befdf4e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4182, + "serialized_chars": 4181, "estimated_input_tokens": 1046, "estimated_eval_tokens": 2348 }, @@ -26954,7 +26908,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4182, + "serialized_chars": 4181, "estimated_input_tokens": 1046, "estimated_eval_tokens": 2348 }, @@ -26963,18 +26917,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are unrelated: one re-enables a VITS forward test and adjusts its assertions, while the other updates an RT-DETR v2 weight-conversion script for model key changes. They do not describe the same bug or fix path.", - "confidence": 0.98, + "summary": "The two PRs are unrelated: one re-enables a VITS model forward test, while the other adjusts an RT-DETR v2 weight-conversion script for current model keys. They do not fix the same bug or change the same concrete code path.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: PR #47105 is a VITS test update, while PR #47129 is an RT-DETR v2 converter fix; they target different subsystems and problems.", "best_issue_reason": null, - "best_pr_reason": "Neither PR is a good cluster representative because they target different models and different kinds of changes (test behavior vs. conversion mapping).", + "best_pr_reason": "No single best PR for deduplication. Both are valid standalone changes, but neither is a duplicate of the other or a better representative for the pair.", "soft_edge_verdicts": [ { "left": "pull_request:47105", "right": "pull_request:47129", "accept": false, - "reason": "Different subsystems and different code paths: VITS test re-enable/assertion cleanup vs. RT-DETR v2 converter key-mapping update. They are not the same underlying change and would not plausibly be merged into one PR." + "reason": "Different models and code paths: VITS test re-enable vs RT-DETR v2 converter key mapping fix." } ] }, @@ -26996,92 +26950,110 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-47122-4", + "cluster_id": "cluster-47122-5", "nodes": [ "pull_request:47122", "pull_request:47123", "pull_request:47124", - "pull_request:47150" + "pull_request:47150", + "pull_request:47224" ], "soft_pairs": [ "pull_request:47122|pull_request:47123", + "pull_request:47122|pull_request:47224", "pull_request:47122|pull_request:47124", "pull_request:47122|pull_request:47150", "pull_request:47123|pull_request:47124", "pull_request:47123|pull_request:47150", - "pull_request:47124|pull_request:47150" + "pull_request:47124|pull_request:47150", + "pull_request:47150|pull_request:47224" ], - "prepared_review_unit_hash": "b25a63efeb4b3682386039fe888c0e6e786ae706fa7f876d368efb03ade1b096", + "prepared_review_unit_hash": "47e98445d8efee7c841090061eb41bc0366a4a1a767495c63ef5f9a73d29062b", "component_budget": { - "node_count": 4, - "item_count": 4, - "soft_pair_count": 6, - "serialized_chars": 9335, - "estimated_input_tokens": 2334, - "estimated_eval_tokens": 4924 + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 12423, + "estimated_input_tokens": 3106, + "estimated_eval_tokens": 6468 }, "budget": { - "node_count": 4, - "item_count": 4, - "soft_pair_count": 6, - "serialized_chars": 9335, - "estimated_input_tokens": 2334, - "estimated_eval_tokens": 4924 + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 12423, + "estimated_input_tokens": 3106, + "estimated_eval_tokens": 6468 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four PRs are unrelated model-specific fixes: EdgeTam runtime kwargs handling, MiniMax-M3-VL test loading/offload setup, Falcon-Mamba test expectation updates, and Whisper generation length adjustment. They do not look like duplicates of the same underlying bug or change.", - "confidence": 0.98, + "summary": "This cluster is mostly a set of unrelated PRs for different model/test failures, with one umbrella PR (47224) that appears to subsume at least the EdgeTam and Whisper fixes.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: these are different code paths and failure modes across different models, so there is no duplicate cluster to collapse into one PR.", + "canonical_pr_reason": "PR 47224 is the broad umbrella change that bundles multiple Serge integration-test fixes and appears to include the concrete EdgeTam and Whisper code-path corrections from smaller PRs.", "best_issue_reason": null, - "best_pr_reason": "No global best PR for deduplication purposes; each PR addresses a distinct model-specific problem, so none is a suitable representative of the others.", + "best_pr_reason": "47224 is the best cluster representative because it is the most comprehensive fix and explicitly targets the shared integration-test-failure cleanup, while the other PRs are narrower point fixes.", "soft_edge_verdicts": [ { "left": "pull_request:47122", "right": "pull_request:47123", "accept": false, - "reason": "Different failures and fixes: EdgeTam changes model forward kwargs handling, while MiniMax-M3-VL only adjusts a test loader with an offload folder. Same general area of integration tests, but not the same bug or mergeable code change." + "reason": "Different failures and code paths: EdgeTam forward kwargs handling vs. MiniMax-M3-VL load/offload test setup." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47224", + "accept": true, + "reason": "47224 includes the same EdgeTam fix theme in the same model files (dropping unsupported original_sizes/backbone kwargs), so it looks like the same underlying change." }, { "left": "pull_request:47122", "right": "pull_request:47124", "accept": false, - "reason": "EdgeTam runtime fix vs Falcon-Mamba expected-output test updates; they touch different models and unrelated behaviors." + "reason": "EdgeTam kwargs filtering is unrelated to Falcon-Mamba generation expectation updates." }, { "left": "pull_request:47122", "right": "pull_request:47150", "accept": false, - "reason": "EdgeTam drops unsupported processor kwargs; Whisper changes generation max_length logic. Completely different code paths and symptoms." + "reason": "EdgeTam backbone kwarg cleanup and Whisper max-length handling are separate bugs." }, { "left": "pull_request:47123", "right": "pull_request:47124", "accept": false, - "reason": "MiniMax-M3-VL load-error test setup fix vs Falcon-Mamba output-mismatch expectation updates. Different models and distinct failure modes." + "reason": "MiniMax-M3-VL offload-folder test fix is unrelated to Falcon-Mamba output mismatches." }, { "left": "pull_request:47123", "right": "pull_request:47150", "accept": false, - "reason": "MiniMax-M3-VL test loading/offload issue is unrelated to Whisper generation length handling." + "reason": "MiniMax-M3-VL loading issue and Whisper generation-length issue are different problems." }, { "left": "pull_request:47124", "right": "pull_request:47150", "accept": false, - "reason": "Falcon-Mamba test expectation updates are not the same underlying bug as Whisper's generation_config max_length correction." + "reason": "Both are test-mismatch PRs, but they fix different models and different failure modes." + }, + { + "left": "pull_request:47150", + "right": "pull_request:47224", + "accept": true, + "reason": "47224 appears to include the same Whisper generation_whisper.py correction for max_length on previous-segment conditioning, so this is the same underlying fix." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The representative PR choice is supported by the umbrella nature of 47224, and the accepted soft-edge pairs (47122\u219447224, 47150\u219447224) are plausibly mergeable as the same bugfix themes. The rejected pairs are also reasonably treated as unrelated point fixes across different models/failures." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null } diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index 726daad905a526d601c9164d5a5e95c61317c079..dc46cb29261e91a5eb11e6a61ca80ad3bd5fa202 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260709t180026z", + "analysis_id": "hybrid-model-20260710t000031z", "archived_artifacts": { - "hybrid": "snapshots/20260709T180026Z/analysis-runs/hybrid-model-20260709t180026z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260709T180026Z/analysis-runs/hybrid-model-20260709t180026z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-07-09T18:15:08Z", + "published_at": "2026-07-10T00:11:43Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260709T180026Z", + "snapshot_id": "20260710T000031Z", "variant": "hybrid" } diff --git a/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..4d857172665091410d9538a67887525705419cbc --- /dev/null +++ b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json @@ -0,0 +1,4321 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260710T000031Z", + "generated_at": "2026-07-10T00:11:40Z", + "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-46376-5", + "summary": "Cluster of 5 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46836, + "issue_numbers": [], + "pr_numbers": [ + 46376, + 46509, + 46836, + 47143, + 47163 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46376, + "right_pr_number": 46509, + "code_similarity": 0.152, + "size_similarity": 0.627, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.176, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 46836, + "code_similarity": 0.071, + "size_similarity": 0.224, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.176, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 47143, + "code_similarity": 0.104, + "size_similarity": 0.431, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.121, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46376, + "right_pr_number": 47163, + "code_similarity": 0.009, + "size_similarity": 0.022, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.028, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 46836, + "code_similarity": 0.119, + "size_similarity": 0.358, + "file_overlap": 0.056, + "area_overlap": 0.0, + "patch_similarity": 0.203, + "shared_filenames": [ + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 412, + 418 + ], + [ + 1071, + 1077 + ], + [ + 1199, + 1205 + ] + ], + "right_ranges": [ + [ + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 1092, + 1098 + ] + ], + "right_ranges": [ + [ + 509, + 515 + ], + [ + 1119, + 1125 + ] + ] + } + ] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47143, + "code_similarity": 0.235, + "size_similarity": 0.687, + "file_overlap": 0.091, + "area_overlap": 0.143, + "patch_similarity": 0.135, + "shared_filenames": [ + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 412, + 418 + ], + [ + 1071, + 1077 + ], + [ + 1199, + 1205 + ] + ], + "right_ranges": [ + [ + 412, + 418 + ], + [ + 975, + 981 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 1092, + 1098 + ] + ], + "right_ranges": [ + [ + 348, + 354 + ], + [ + 1960, + 1966 + ] + ] + } + ] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47163, + "code_similarity": 0.006, + "size_similarity": 0.014, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.024, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46836, + "right_pr_number": 47143, + "code_similarity": 0.156, + "size_similarity": 0.52, + "file_overlap": 0.077, + "area_overlap": 0.0, + "patch_similarity": 0.191, + "shared_filenames": [ + "src/transformers/models/__init__.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/__init__.py", + "left_ranges": [ + [ + 453, + 459 + ] + ], + "right_ranges": [ + [ + 305, + 311 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 + ] + ], + "right_ranges": [ + [ + 412, + 418 + ], + [ + 975, + 981 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 509, + 515 + ], + [ + 1119, + 1125 + ] + ], + "right_ranges": [ + [ + 348, + 354 + ], + [ + 1960, + 1966 + ] + ] + } + ] + }, + { + "left_pr_number": 46836, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.005, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.021, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47143, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.01, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.013, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46894-5", + "summary": "Cluster of 1 issues and 4 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46894, + "canonical_pr_number": 46903, + "issue_numbers": [ + 46894 + ], + "pr_numbers": [ + 46895, + 46903, + 46958, + 47090 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46895, + "right_pr_number": 46903, + "code_similarity": 0.533, + "size_similarity": 0.404, + "file_overlap": 0.333, + "area_overlap": 0.933, + "patch_similarity": 0.17, + "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, + 650 + ] + ], + "right_ranges": [ + [ + 637, + 651 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 513 + ] + ], + "right_ranges": [ + [ + 500, + 514 + ] + ] + } + ] + }, + { + "left_pr_number": 46895, + "right_pr_number": 46958, + "code_similarity": 0.779, + "size_similarity": 0.915, + "file_overlap": 1.0, + "area_overlap": 0.5, + "patch_similarity": 0.81, + "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": [ + [ + 637, + 650 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 513 + ] + ], + "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": 46895, + "right_pr_number": 47090, + "code_similarity": 0.819, + "size_similarity": 0.932, + "file_overlap": 1.0, + "area_overlap": 0.783, + "patch_similarity": 0.389, + "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": [ + [ + 637, + 650 + ] + ], + "right_ranges": [ + [ + 637, + 649 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 513 + ] + ], + "right_ranges": [ + [ + 500, + 512 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 35, + 41 + ], + [ + 176, + 214 + ] + ], + "right_ranges": [ + [ + 175, + 222 + ] + ] + } + ] + }, + { + "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 + ] + ] + } + ] + }, + { + "left_pr_number": 46903, + "right_pr_number": 47090, + "code_similarity": 0.519, + "size_similarity": 0.433, + "file_overlap": 0.333, + "area_overlap": 0.867, + "patch_similarity": 0.192, + "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": [ + [ + 637, + 649 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 500, + 514 + ] + ], + "right_ranges": [ + [ + 500, + 512 + ] + ] + } + ] + }, + { + "left_pr_number": 46958, + "right_pr_number": 47090, + "code_similarity": 0.742, + "size_similarity": 0.852, + "file_overlap": 1.0, + "area_overlap": 0.611, + "patch_similarity": 0.387, + "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": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ], + "right_ranges": [ + [ + 637, + 649 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ], + "right_ranges": [ + [ + 500, + 512 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 15, + 21 + ], + [ + 180, + 221 + ] + ], + "right_ranges": [ + [ + 175, + 222 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46693-4", + "summary": "Cluster of 1 issues and 3 PRs centered on issue #46693.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46693, + "canonical_pr_number": 47134, + "issue_numbers": [ + 46693 + ], + "pr_numbers": [ + 46696, + 47134, + 47215 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46696, + "right_pr_number": 47134, + "code_similarity": 0.464, + "size_similarity": 0.8, + "file_overlap": 1.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [ + "src/transformers/modeling_flash_attention_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/modeling_flash_attention_utils.py", + "left_ranges": [ + [ + 667, + 692 + ] + ], + "right_ranges": [ + [ + 316, + 323 + ], + [ + 366, + 373 + ] + ] + } + ] + }, + { + "left_pr_number": 46696, + "right_pr_number": 47215, + "code_similarity": 0.138, + "size_similarity": 0.022, + "file_overlap": 0.2, + "area_overlap": 0.191, + "patch_similarity": 0.043, + "shared_filenames": [ + "src/transformers/modeling_flash_attention_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/modeling_flash_attention_utils.py", + "left_ranges": [ + [ + 667, + 692 + ] + ], + "right_ranges": [ + [ + 33, + 39 + ], + [ + 317, + 324 + ], + [ + 368, + 375 + ], + [ + 491, + 499 + ], + [ + 680, + 702 + ] + ] + } + ] + }, + { + "left_pr_number": 47134, + "right_pr_number": 47215, + "code_similarity": 0.144, + "size_similarity": 0.018, + "file_overlap": 0.2, + "area_overlap": 0.224, + "patch_similarity": 0.015, + "shared_filenames": [ + "src/transformers/modeling_flash_attention_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/modeling_flash_attention_utils.py", + "left_ranges": [ + [ + 316, + 323 + ], + [ + 366, + 373 + ] + ], + "right_ranges": [ + [ + 33, + 39 + ], + [ + 317, + 324 + ], + [ + 368, + 375 + ], + [ + 491, + 499 + ], + [ + 680, + 702 + ] + ] + } + ] + } + ] + }, + { + "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-46821-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46821.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 46821, + "canonical_pr_number": 46834, + "issue_numbers": [ + 46821 + ], + "pr_numbers": [ + 46834, + 47223 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 46834, + "right_pr_number": 47223, + "code_similarity": 0.242, + "size_similarity": 0.644, + "file_overlap": 0.333, + "area_overlap": 0.0, + "patch_similarity": 0.088, + "shared_filenames": [ + "src/transformers/models/timesfm2_5/modular_timesfm2_5.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/timesfm2_5/modular_timesfm2_5.py", + "left_ranges": [ + [ + 564, + 570 + ], + [ + 651, + 671 + ] + ], + "right_ranges": [ + [ + 101, + 106 + ] + ] + } + ] + } + ] + }, + { + "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": 47162, + "issue_numbers": [ + 46305 + ], + "pr_numbers": [ + 46310, + 47162 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46310, + "right_pr_number": 47162, + "code_similarity": 0.577, + "size_similarity": 0.338, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.062, + "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-47205-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #47205.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 47205, + "canonical_pr_number": 47218, + "issue_numbers": [ + 47205 + ], + "pr_numbers": [ + 47218, + 47219 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47218, + "right_pr_number": 47219, + "code_similarity": 0.635, + "size_similarity": 0.755, + "file_overlap": 0.909, + "area_overlap": 0.46, + "patch_similarity": 0.333, + "shared_filenames": [ + "tests/models/aimv2/test_modeling_aimv2.py", + "tests/models/aria/test_modeling_aria.py", + "tests/models/deepseek_vl_hybrid/test_image_processing_deepseek_vl_hybrid.py", + "tests/models/eomt/test_modeling_eomt.py", + "tests/models/eomt_dinov3/test_modeling_eomt_dinov3.py", + "tests/models/fast_vlm/test_modeling_fast_vlm.py", + "tests/models/flava/test_image_processing_flava.py", + "tests/models/fuyu/test_image_processing_fuyu.py", + "tests/models/idefics/test_image_processing_idefics.py", + "tests/models/imagegpt/test_image_processing_imagegpt.py", + "tests/models/kimi_k25/test_image_processing_kimi_k25.py", + "tests/models/layoutlmv2/test_image_processing_layoutlmv2.py", + "tests/models/lfm2_vl/test_modeling_lfm2_vl.py", + "tests/models/llava/test_modeling_llava.py", + "tests/models/minimax_m3_vl/test_modeling_minimax_m3_vl.py", + "tests/models/pix2struct/test_image_processing_pix2struct.py", + "tests/models/qwen2_vl/test_image_processing_qwen2_vl.py", + "tests/models/video_llama_3/test_image_processing_video_llama_3.py", + "tests/models/vitpose/test_image_processing_vitpose.py", + "tests/test_image_processing_common.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/aimv2/test_modeling_aimv2.py", + "left_ranges": [ + [ + 18, + 23 + ], + [ + 34, + 40 + ], + [ + 58, + 63 + ], + [ + 477, + 485 + ], + [ + 513, + 521 + ], + [ + 546, + 554 + ] + ], + "right_ranges": [ + [ + 18, + 23 + ], + [ + 43, + 49 + ], + [ + 58, + 63 + ], + [ + 478, + 484 + ], + [ + 512, + 518 + ], + [ + 543, + 549 + ] + ] + }, + { + "filename": "tests/models/aria/test_modeling_aria.py", + "left_ranges": [ + [ + 16, + 21 + ], + [ + 42, + 48 + ], + [ + 51, + 57 + ], + [ + 246, + 252 + ], + [ + 286, + 292 + ], + [ + 320, + 329 + ], + [ + 360, + 369 + ], + [ + 402, + 411 + ], + [ + 437, + 444 + ] + ], + "right_ranges": [ + [ + 44, + 50 + ], + [ + 322, + 328 + ], + [ + 360, + 366 + ], + [ + 400, + 406 + ] + ] + }, + { + "filename": "tests/models/deepseek_vl_hybrid/test_image_processing_deepseek_vl_hybrid.py", + "left_ranges": [ + [ + 19, + 25 + ], + [ + 203, + 210 + ] + ], + "right_ranges": [ + [ + 20, + 26 + ], + [ + 204, + 212 + ] + ] + }, + { + "filename": "tests/models/eomt/test_modeling_eomt.py", + "left_ranges": [ + [ + 15, + 26 + ], + [ + 30, + 36 + ], + [ + 174, + 182 + ], + [ + 219, + 227 + ], + [ + 237, + 245 + ], + [ + 276, + 284 + ], + [ + 327, + 335 + ], + [ + 375, + 383 + ] + ], + "right_ranges": [ + [ + 15, + 36 + ], + [ + 174, + 180 + ], + [ + 217, + 223 + ], + [ + 233, + 239 + ], + [ + 270, + 276 + ], + [ + 319, + 325 + ], + [ + 365, + 371 + ] + ] + }, + { + "filename": "tests/models/eomt_dinov3/test_modeling_eomt_dinov3.py", + "left_ranges": [ + [ + 15, + 20 + ], + [ + 27, + 33 + ], + [ + 37, + 43 + ], + [ + 219, + 227 + ], + [ + 263, + 271 + ], + [ + 322, + 330 + ], + [ + 380, + 388 + ], + [ + 428, + 436 + ], + [ + 472, + 480 + ] + ], + "right_ranges": [ + [ + 15, + 20 + ], + [ + 29, + 43 + ], + [ + 219, + 225 + ], + [ + 261, + 267 + ], + [ + 318, + 324 + ], + [ + 374, + 380 + ], + [ + 420, + 426 + ], + [ + 462, + 468 + ] + ] + }, + { + "filename": "tests/models/fast_vlm/test_modeling_fast_vlm.py", + "left_ranges": [ + [ + 16, + 21 + ], + [ + 36, + 42 + ], + [ + 45, + 51 + ], + [ + 256, + 262 + ], + [ + 280, + 289 + ] + ], + "right_ranges": [ + [ + 39, + 45 + ], + [ + 283, + 289 + ] + ] + }, + { + "filename": "tests/models/flava/test_image_processing_flava.py", + "left_ranges": [ + [ + 12, + 26 + ], + [ + 401, + 408 + ] + ], + "right_ranges": [ + [ + 12, + 27 + ], + [ + 402, + 408 + ] + ] + }, + { + "filename": "tests/models/fuyu/test_image_processing_fuyu.py", + "left_ranges": [ + [ + 1, + 5 + ], + [ + 14, + 20 + ], + [ + 178, + 185 + ] + ], + "right_ranges": [ + [ + 1, + 5 + ], + [ + 15, + 21 + ], + [ + 179, + 185 + ] + ] + }, + { + "filename": "tests/models/idefics/test_image_processing_idefics.py", + "left_ranges": [ + [ + 11, + 18 + ], + [ + 25, + 31 + ], + [ + 149, + 156 + ] + ], + "right_ranges": [ + [ + 11, + 18 + ], + [ + 26, + 32 + ], + [ + 150, + 156 + ] + ] + }, + { + "filename": "tests/models/imagegpt/test_image_processing_imagegpt.py", + "left_ranges": [ + [ + 20, + 25 + ], + [ + 33, + 39 + ], + [ + 284, + 291 + ] + ], + "right_ranges": [ + [ + 20, + 25 + ], + [ + 34, + 40 + ], + [ + 285, + 291 + ] + ] + }, + { + "filename": "tests/models/kimi_k25/test_image_processing_kimi_k25.py", + "left_ranges": [ + [ + 12, + 31 + ], + [ + 270, + 277 + ] + ], + "right_ranges": [ + [ + 12, + 27 + ], + [ + 266, + 272 + ] + ] + }, + { + "filename": "tests/models/layoutlmv2/test_image_processing_layoutlmv2.py", + "left_ranges": [ + [ + 15, + 20 + ], + [ + 29, + 42 + ], + [ + 147, + 154 + ] + ], + "right_ranges": [ + [ + 15, + 20 + ], + [ + 30, + 43 + ], + [ + 148, + 154 + ] + ] + }, + { + "filename": "tests/models/lfm2_vl/test_modeling_lfm2_vl.py", + "left_ranges": [ + [ + 15, + 22 + ], + [ + 34, + 45 + ], + [ + 209, + 219 + ], + [ + 291, + 301 + ] + ], + "right_ranges": [ + [ + 37, + 43 + ], + [ + 211, + 217 + ], + [ + 295, + 301 + ] + ] + }, + { + "filename": "tests/models/llava/test_modeling_llava.py", + "left_ranges": [ + [ + 17, + 22 + ], + [ + 42, + 48 + ], + [ + 52, + 58 + ], + [ + 305, + 311 + ], + [ + 334, + 340 + ], + [ + 366, + 375 + ], + [ + 419, + 428 + ], + [ + 474, + 483 + ], + [ + 529, + 536 + ], + [ + 610, + 616 + ], + [ + 637, + 644 + ], + [ + 666, + 673 + ], + [ + 699, + 706 + ] + ], + "right_ranges": [ + [ + 45, + 51 + ], + [ + 368, + 374 + ], + [ + 419, + 425 + ], + [ + 472, + 478 + ] + ] + }, + { + "filename": "tests/models/minimax_m3_vl/test_modeling_minimax_m3_vl.py", + "left_ranges": [ + [ + 37, + 43 + ], + [ + 652, + 657 + ], + [ + 695, + 703 + ] + ], + "right_ranges": [ + [ + 45, + 51 + ], + [ + 697, + 703 + ] + ] + }, + { + "filename": "tests/models/pix2struct/test_image_processing_pix2struct.py", + "left_ranges": [ + [ + 21, + 27 + ], + [ + 99, + 106 + ] + ], + "right_ranges": [ + [ + 22, + 28 + ], + [ + 99, + 107 + ] + ] + }, + { + "filename": "tests/models/qwen2_vl/test_image_processing_qwen2_vl.py", + "left_ranges": [ + [ + 12, + 35 + ], + [ + 297, + 304 + ] + ], + "right_ranges": [ + [ + 12, + 22 + ], + [ + 25, + 31 + ], + [ + 293, + 299 + ] + ] + }, + { + "filename": "tests/models/video_llama_3/test_image_processing_video_llama_3.py", + "left_ranges": [ + [ + 18, + 35 + ], + [ + 300, + 307 + ] + ], + "right_ranges": [ + [ + 18, + 31 + ], + [ + 296, + 302 + ] + ] + }, + { + "filename": "tests/models/vitpose/test_image_processing_vitpose.py", + "left_ranges": [ + [ + 11, + 25 + ], + [ + 231, + 238 + ] + ], + "right_ranges": [ + [ + 11, + 26 + ], + [ + 232, + 238 + ] + ] + }, + { + "filename": "tests/test_image_processing_common.py", + "left_ranges": [ + [ + 12, + 17 + ], + [ + 20, + 25 + ], + [ + 47, + 73 + ], + [ + 194, + 200 + ] + ], + "right_ranges": [ + [ + 12, + 17 + ], + [ + 20, + 28 + ], + [ + 174, + 180 + ] + ] + } + ] + } + ] + }, + { + "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-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-47122-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47122, + "issue_numbers": [], + "pr_numbers": [ + 47122, + 47150, + 47224 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47122, + "right_pr_number": 47150, + "code_similarity": 0.085, + "size_similarity": 0.423, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47224, + "code_similarity": 0.474, + "size_similarity": 0.302, + "file_overlap": 0.4, + "area_overlap": 0.818, + "patch_similarity": 0.051, + "shared_filenames": [ + "src/transformers/models/edgetam/modeling_edgetam.py", + "src/transformers/models/edgetam/modular_edgetam.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/edgetam/modeling_edgetam.py", + "left_ranges": [ + [ + 464, + 472 + ] + ], + "right_ranges": [ + [ + 464, + 474 + ] + ] + }, + { + "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "left_ranges": [ + [ + 210, + 218 + ] + ], + "right_ranges": [ + [ + 210, + 220 + ] + ] + } + ] + }, + { + "left_pr_number": 47150, + "right_pr_number": 47224, + "code_similarity": 0.442, + "size_similarity": 0.128, + "file_overlap": 0.2, + "area_overlap": 1.0, + "patch_similarity": 0.04, + "shared_filenames": [ + "src/transformers/models/whisper/generation_whisper.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/whisper/generation_whisper.py", + "left_ranges": [ + [ + 1937, + 1943 + ] + ], + "right_ranges": [ + [ + 1937, + 1943 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-47077-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47077, + "issue_numbers": [], + "pr_numbers": [ + 47077, + 47181 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47077, + "right_pr_number": 47181, + "code_similarity": 0.029, + "size_similarity": 0.086, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.077, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-47034-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47216, + "issue_numbers": [], + "pr_numbers": [ + 47034, + 47216 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47034, + "right_pr_number": 47216, + "code_similarity": 0.214, + "size_similarity": 0.982, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.117, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-44001-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44004, + "issue_numbers": [], + "pr_numbers": [ + 44001, + 44004 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44001, + "right_pr_number": 44004, + "code_similarity": 0.087, + "size_similarity": 0.237, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.262, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-44002-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44002, + "issue_numbers": [], + "pr_numbers": [ + 44002, + 44029 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44002, + "right_pr_number": 44029, + "code_similarity": 0.135, + "size_similarity": 0.43, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.329, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [], + "duplicate_prs": [ + { + "cluster_id": "cluster-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-44001-2", + "canonical_pr_number": 44004, + "duplicate_pr_numbers": [ + 44001 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44001-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-44002-2", + "canonical_pr_number": 44002, + "duplicate_pr_numbers": [ + 44029 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44002-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45561-3", + "canonical_pr_number": 45645, + "duplicate_pr_numbers": [ + 45639 + ], + "target_issue_number": 45561, + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45647-3", + "canonical_pr_number": 45738, + "duplicate_pr_numbers": [ + 45737 + ], + "target_issue_number": 45647, + "reason": "PRs in cluster-45647-3 are treated as duplicates because they converge on issue #45647 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-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": 47162, + "duplicate_pr_numbers": [ + 46310 + ], + "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-46376-5", + "canonical_pr_number": 46836, + "duplicate_pr_numbers": [ + 46376, + 46509, + 47143, + 47163 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46376-5 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46693-4", + "canonical_pr_number": 47134, + "duplicate_pr_numbers": [ + 46696, + 47215 + ], + "target_issue_number": 46693, + "reason": "PRs in cluster-46693-4 are treated as duplicates because they converge on issue #46693 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46821-3", + "canonical_pr_number": 46834, + "duplicate_pr_numbers": [ + 47223 + ], + "target_issue_number": 46821, + "reason": "PRs in cluster-46821-3 are treated as duplicates because they converge on issue #46821 with closing_reference 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-46894-5", + "canonical_pr_number": 46903, + "duplicate_pr_numbers": [ + 46895, + 46958, + 47090 + ], + "target_issue_number": 46894, + "reason": "PRs in cluster-46894-5 are treated as duplicates because they converge on issue #46894 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47034-2", + "canonical_pr_number": 47216, + "duplicate_pr_numbers": [ + 47034 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47034-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47077-2", + "canonical_pr_number": 47077, + "duplicate_pr_numbers": [ + 47181 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47077-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47122-3", + "canonical_pr_number": 47122, + "duplicate_pr_numbers": [ + 47150, + 47224 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47122-3 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47205-3", + "canonical_pr_number": 47218, + "duplicate_pr_numbers": [ + 47219 + ], + "target_issue_number": 47205, + "reason": "PRs in cluster-47205-3 are treated as duplicates because they converge on issue #47205 with closing_reference, shared_issue_target, soft_similarity evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-46894-5", + "issue_number": 46894, + "reason": "Issue #46894 is the strongest global issue candidate because it is open, belongs to a cluster with 5 artifacts, and carries 1 discussion comments plus 9 inbound references.", + "score": 161.0 + }, + "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/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..42e43c41f066a512fc4fe4b04fdccba9ee0737a4 --- /dev/null +++ b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,27061 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260710T000031Z", + "generated_at": "2026-07-10T00:11:40Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:46165|issue:46468", + "issue:44805|issue:45161", + "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:24643|issue:30277", + "issue:43716|issue:46869", + "issue:34928|issue:44805", + "issue:45685|issue:46869", + "issue:45859|issue:45902", + "issue:34928|issue:36331", + "issue:43638|issue:45161", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:44292|issue:45237", + "issue:45874|issue:46294", + "issue:39290|issue:45081", + "issue:45800|issue:46515", + "issue:45823|issue:45874", + "issue:34928|issue:41720", + "issue:43425|issue:44292", + "issue:43828|issue:45161", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:45685", + "issue:45507|issue:45685", + "issue:44292|issue:44918", + "issue:43716|issue:44292", + "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": "9108cae3644037c8968aa5d06377e9235abba0a236f1b2a1f1b5b35e335f6785", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 62, + "serialized_chars": 46276, + "estimated_input_tokens": 11569, + "estimated_eval_tokens": 23394 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous, with only a few true duplicate subgroups: ZeRO-3 load-time index errors, Qwen3.5Moe rotary/offload issues, and a torch 2.4.1 compatibility report. Most other soft pairs are same subsystem but different concrete failures, models, or backends.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43638 is the best representative of the strongest true duplicate family: it describes the shared ZeRO-3 load-time `index 0 is out of bounds for dimension 0 with size 0` failure, which is also reported in issue:43596 and issue:41762.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43638 is the clearest anchor for the cluster\u2019s only well-supported duplicate subgroup, and its title is general enough to represent that specific Zero-3 empty-tensor indexing bug better than the more model-specific variants.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failures: config attribute error vs streaming chat/completions API error on a different model." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different backend/model problems: AMD GPU failure vs MPS histogram op not implemented for Int." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Both are device/backend-related, but the concrete bugs differ: dtype mismatch in multimodal preprocessing vs AMD GPU runtime failure." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Different concrete dtype-mismatch paths; same broad theme is not enough." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Different models and failure modes; one is an AMD GPU runtime issue, the other a broken example." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific failures: AMD GPU support vs invalid method call in GraniteMoEHybrid." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "Different code paths: DeepSpeed collective mismatch during Jamba training vs ONNX export failure for ModernBERT." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "Different issues in different models; ONNX export bug vs Qwen3Moe FineGrainedFP8 failure." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Different concrete failures: export-time issue vs cudaErrorAssert during auto device mapping." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Docs/example quality issue vs runtime task crash; not the same underlying bug." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "Unrelated failures: export error vs import-time GenerationMixin issue." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are stale v5 docs, but they reference different removed content and don't look like one concrete duplicate change." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "Different training failures: Qwen3.5 input embedding unpacking vs resume/trust-remote-code behavior." + }, + { + "left": "issue:46165", + "right": "issue:46869", + "accept": false, + "reason": "Different models and symptoms: resume/trust-remote-code issue vs multi-GPU YOLO device-map failure." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Both mention device mismatch, but one is a mask/indexing shape error and the other is a forward device-map mismatch on Gemma4 PLE." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Same general training/runtime area, but the failures are different: ZeRO-3 zero-sized tensor indexing vs mask/tensor shape mismatch." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "Different bugs: ZeRO-3 initialization index error vs resume/trust-remote-code training behavior." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Different failure modes; both are errors, but only one is the Zero-3 empty-tensor indexing bug." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal/video chat-template bug vs tensor mask shape mismatch." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Different topics: model resume/trust-remote-code vs broken official example." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Different concrete problems: mask/index shape mismatch vs GPT-OSS tensor-parallel path not working." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Different bugs in different areas; video StopIteration vs training resume issue." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "Different model-specific failures: BarkProcessor voice preset vs ModernBERT ONNX export." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Both involve distributed training, but the actual bugs differ: collective mismatch vs Zero-3 zero-sized tensor indexing." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same Zero-3 load-time `index 0 is out of bounds for dimension 0 with size 0` failure; the model wording differs but the underlying bug and code path match." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Different issues: Liger Kernel cross-entropy routing vs GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Different failures: v5 detokenization regression vs LoRA merge/save/reload inconsistency." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Different models and bugs; NVFP4 runtime error is unrelated to LoRA merge behavior." + }, + { + "left": "issue:44841", + "right": "issue:46661", + "accept": false, + "reason": "Processor failure for Voxtral vs remote-code model_class/config_class attribute error." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "Both are DeepSpeed training issues, but one is a 2-D weight assertion and the other is a collective mismatch on Jamba/ZeRO-3." + }, + { + "left": "issue:43716", + "right": "issue:46869", + "accept": false, + "reason": "Different models and concrete failures; dtype mismatch in Mistral-3 vs YOLO multi-GPU device-map failure." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Both are shape/runtime errors under distributed training, but they are different bugs and code paths." + }, + { + "left": "issue:45685", + "right": "issue:46869", + "accept": false, + "reason": "Different backend/model failures: MPS histogram kernel gap vs YOLO multi-GPU device-map error." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": true, + "reason": "Same Qwen3_5Moe rotary embedding offload/materialization bug: one reports CPU offload incompatibility, the other the uninitialized `inv_freq` symptom after `to_empty(cuda)`." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "Different training bugs: activation-checkpointing/FSDP recomputation mismatch vs `compute_loss` signature change." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "Different code paths: ZeRO-3 index error vs GPT-OSS tensor-parallel issue." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different concrete failures; same broad distributed-training label is not enough." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific failures: dtype mismatch in Mistral-3 vs GraniteMoEHybrid invalid method call." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Different hardware/model bugs; both are runtime failures but not the same underlying issue." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal/model behaviors: expensive reverse embedding lookup vs video feature/token mismatch." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Unrelated regressions: Gemma3TextConfig attribute error vs Mistral regex patch crash." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": true, + "reason": "Same torch 2.4.1 compatibility report; the second issue is just a shorter restatement of the same incompatibility." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Different Gemma4 problems: device mismatch during forward vs reverse embedding lookup cost." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Different version/model failures; generic torch incompatibility vs Qwen-3-8B-NVFP4 runtime error." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Both involve model execution, but one is autocast dtype mismatch and the other is tensor-parallel MoE failure." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "Different issues: grouped_mm tracing error in Mixtral vs dtype mismatch in Mistral-3 preprocessing." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Different models and bugs; dtype mismatch under autocast is unrelated to NVFP4 runtime failure." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Different hardware/backend issues with different model contexts; not the same bug." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Different model/backend failures: invalid method call vs unimplemented MPS histogram op." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Different failures: NVFP4 runtime issue vs TRL SFT unpacking problem." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different model/runtime errors; both are about model execution but not the same underlying bug." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "Different training paths and failure modes: TRL SFT input unpacking vs GPT-OSS tensor-parallel issue." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Different areas: Liger Kernel cross-entropy call vs Qwen3.5 embedding unpacking in TRL SFT." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 `index 0 is out of bounds for dimension 0 with size 0` load-time failure; the model names differ, but the bug pattern matches." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are stale docs issues, but they cover different removed functionality and look like separate cleanup items." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Different concrete issues: autocast dtype mismatch vs MPS kernel not implemented." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Different bugs: Liger Kernel cross-entropy routing vs mask/tensor indexing shape error." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the accepted soft-edge pairs are all based on close title-level matches rather than broad thematic similarity. The conservative rejections are also appropriate. The only mild concern is wording like \"only well-supported duplicate subgroup,\" but the rest of the summary correctly acknowledges multiple small duplicate families, so overall this is acceptable." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 2, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:39290|issue:41720", + "issue:41762|issue:43596", + "issue:43901|issue:44512", + "issue:45237|issue:45901", + "issue:44589|issue:46304", + "issue:43039|issue:44560", + "issue:45823|issue:46294", + "issue:44841|issue:45823", + "issue:43975|issue:46015", + "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:43792|issue:45070", + "issue:43975|issue:45901", + "issue:45362|issue:45464", + "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:45507|issue:46869", + "issue:43404|issue:43866", + "issue:41093|issue:41762", + "issue:45399|issue:46661", + "issue:43572|issue:45070", + "issue:45800|issue:45823" + ], + "prepared_review_unit_hash": "c77ae1bc2e9fd62a3e4c89eab6deb7bb32d78061f0707500d64ba55adb25c5a7", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 40047, + "estimated_input_tokens": 10012, + "estimated_eval_tokens": 20280 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct model- or subsystem-specific bugs. The strongest duplicate signal is the v5 AutoTokenizer regression affecting multiple model families; the rest are too loosely related to merge safely.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45920 is the best representative because it describes the tokenizer regression broadly across multiple model families and clearly frames the underlying v5 token-ID bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the most actionable and general issue in the set, with the broadest scope for a single fix and the clearest regression characterization.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Both mention torch/transformers compatibility, but they are different version breakages with different likely causes and no shared concrete code path." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Different failures: vLLM model loading on transformers 4.53.0 versus PyTorch support broken in 5.9.0. Same ecosystem, not the same bug." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "One is a corrupted checkpoint/model loading problem, the other is missing text encoder weights in Sam3Model. Not the same underlying issue." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Both are runtime errors in model execution, but they involve different models, different configs, and different failure modes." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Both involve ZeRO-3 and IndexError, but the model types and reported symptoms differ enough that a shared root cause is not established." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs issues, but they cover different outdated references and do not describe the same change." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Different task and failure class; one is an AMD GPU/model runtime issue, the other is a table-question-answering crash." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Unrelated errors: missing Float8 storage versus distributed-build compatibility. No common bug." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger kernel cross-entropy dispatch and Qwen3-VL video StopIteration are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Both are multimodal/model-specific forward-path issues, but one is PLE device mismatch and the other is video token/template mismatch." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Processor failure for Voxtral and device-map mismatch for Gemma4 are different code paths." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "A detokenization regression and outdated RAG examples are unrelated." + }, + { + "left": "issue:43866", + "right": "issue:46772", + "accept": false, + "reason": "Corrupted Ovis2 checkpoint versus ModelOpt/NVFP4 compatibility is not the same bug." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "NemotronH checkpoint loading and RAG docs cleanup are unrelated." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Remote-code model breakage and PyTorch USE_DISTRIBUTED compatibility are different underlying problems." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Both report the same v5 AutoTokenizer regression causing wrong token IDs across models; they look like the same underlying tokenizer bug and could plausibly be fixed in one PR." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error and RAG documentation issues are unrelated." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Checkpoint corruption/load failure versus a model implementation loading issue; no shared fix." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes Int8Params incompatibility and distributed PyTorch support breakage are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but the affected models and concrete errors differ; insufficient evidence of the same bug." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both are ZeRO-3/BERT-related IndexErrors, but one is a mask shape mismatch and the other is an out-of-bounds index; not clearly the same root cause." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper loading failure and table-question-answering crash are unrelated." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Model test loading failure for GLM-4.7-Flash and a corrupted Ovis2 checkpoint are different issues." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch 2.10 incompatibility and a v5.4.0 pydantic config-field regression are not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper can\u2019t run and pydantic config-field breakage are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization regression versus table QA crash are different problems." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Qwen3.5 chat crash and streaming inference failure are separate code paths with different symptoms." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "Both mention IndexError-style failures, but one is a ZeRO-3 BERT initialization problem and the other is a torch/transformers compatibility issue." + }, + { + "left": "issue:43541", + "right": "issue:46869", + "accept": false, + "reason": "Both are model execution problems, but Mixtral torch-dynamo grouped_mm tracing and YOLO multi-GPU device mapping are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Qwen NVFP4 runtime failure and table-question-answering crash do not share an underlying bug." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Gemma3 sliding_window config error and Gemma4 device mismatch are distinct issues." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Streaming chat API failure and MoE MPS histogram/int type error are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin and Gemma3 sliding_window config error are different failures." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Different MoE failures with different operations and symptoms; no shared concrete bug." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Grouped_mm tracing failure and autocast dtype mismatch are separate runtime issues." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Processor failure for Voxtral and template-node compilation error are unrelated." + }, + { + "left": "issue:45507", + "right": "issue:46869", + "accept": false, + "reason": "GraniteMoEHybrid invalid method call and YOLO multi-GPU device-map failure are different model-specific bugs." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Mistral3 lm_head tying bug and corrupted Ovis2 checkpoint are unrelated." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are IndexError reports, but they affect different models and describe different failure shapes; no clear shared root cause." + }, + { + "left": "issue:45399", + "right": "issue:46661", + "accept": false, + "reason": "Flash-attn2 fallback gating and remote-code model_class/config_class handling are different code-path problems." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression and pydantic PretrainedConfig breakage are unrelated." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch/transformers incompatibility and Gemma4 device mismatch are not the same bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The canonical choice and summary are supported by the packet, and the one accepted pair (issue:45812 with issue:45920) is clearly the same v5 AutoTokenizer regression. The other verdicts stay appropriately cautious and do not overstate shared root causes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 3, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:43366|issue:45464", + "issue:44291|issue:45070", + "issue:44387|issue:45005", + "issue:43531|issue:45070", + "issue:41093|issue:43596", + "issue:43716|issue:46294", + "issue:45020|issue:45800", + "issue:44479|issue:46294", + "issue:43054|issue:43994", + "issue:46468|issue:46869", + "issue:43596|issue:46165", + "issue:45759|issue:45902", + "issue:46640|issue:46661", + "issue:43716|issue:44918", + "issue:39186|issue:44918", + "issue:43645|issue:45399", + "issue:40990|issue:44960", + "issue:45632|issue:45698", + "issue:42915|issue:45399", + "issue:43404|issue:45005", + "issue:44360|issue:44512", + "issue:43873|issue:45005", + "issue:44492|issue:44512", + "issue:44530|issue:45005", + "issue:44918|issue:45507", + "issue:45183|issue:46463", + "issue:44863|issue:45005" + ], + "prepared_review_unit_hash": "676b046d88d2edf0fcc986340146de207c0b38b4b33a01848541b3934ed25d01", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36243, + "estimated_input_tokens": 9061, + "estimated_eval_tokens": 18378 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a broad mix of unrelated Transformers/model-serving issues. The pairwise similarities are mostly superficial (shared keywords like dtype, cache, video, or tie weights), so I do not see a trustworthy canonical duplicate or mergeable PR candidate here.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "Different problems: PaddleOCR-VL load failure in vLLM/Transformers versus GGUF architecture support for gpt-oss." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Both mention shape-related failures, but one is a generic mask/tensor IndexError and the other is a Qwen3.5 + TRL SFT embedding unpacking issue." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Different code paths and symptoms: torch dynamo grouped_mm tracing error versus tensor-parallel-only failure on GPT-OSS MoE." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is not the same as Qwen3.5 streaming chat/completions failure." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GGUF support for gpt-oss and Qwen3.5 streaming inference are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "One is an init_empty_weights argument regression; the other is a PretrainedConfig/pydantic field break." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "Int4 quantization OOM is a different bug from tied-weight handling in translation models." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to the pydantic config-field regression." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both are IndexErrors, but the reported conditions differ: mask length mismatch versus zero-sized tensor during deepspeed zero3 init." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Different modalities and failures: image-preprocessor dtype mismatch versus video feature/token count mismatch." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote-code compatibility breakage is not the same as a torch 2.4.1 / transformers 5.8.0 incompatibility." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Both involve video input, but the former is a broad v5.3.0 regression across several models while the latter is a specific Qwen3.5 apply_chat_template mismatch." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but one reports poor embedding quality and the other wrong outputs/loading behavior; not clearly the same bug." + }, + { + "left": "issue:46468", + "right": "issue:46869", + "accept": false, + "reason": "Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "Zero3 init index error with BertModel is unrelated to resuming training a local remote-code model." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Composite-config unwrap logic and uninitialized-memory rotary embedding are separate issues." + }, + { + "left": "issue:46640", + "right": "issue:46661", + "accept": false, + "reason": "FlashAttention 2 issue and remote-code model_class/config_class error are different failure modes." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Image preprocessor dtype mismatch does not match the TRL SFT input-embedding unpacking failure." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP weight dimensionality error and Qwen3.5 TRL unpacking failure are unrelated." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Notebook custom-model initialization regression is unrelated to the flash-attn2 fallback gating bug." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "These are different model-specific correctness issues with no shared code-path indicated." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both touch remote-code/custom-module loading, but one is a cache-path collision and the other is wrong custom module selection after save/load; not clearly the same bug." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn2 fallback checks." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but the affected models and concrete failures differ substantially." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "A DSA indexer ReLU discussion and a docs typo are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is a different problem from tied-weight handling." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "Cache-strategy typo and removed run-command documentation are unrelated docs issues." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and tied-weight translation-model issues are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "Qwen3.5 SFT embedding failure and GraniteMoEHybrid invalid-method call are different code paths." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": false, + "reason": "General input_ids/inputs_embeds XOR validation messaging and T5 decoder-input validation are related only at a high level, not clearly the same bug." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading and tied-weight translation model issues 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": 27, + "review_unit_index": 4, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:43824", + "issue:43828", + "issue:43856", + "issue:43872", + "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:44991", + "issue:45020", + "issue:45081", + "issue:45399", + "issue:45520", + "issue:45663", + "issue:45685", + "issue:45701", + "issue:45823", + "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:39692|issue:43054", + "issue:43950|issue:44534", + "issue:36331|issue:41762", + "issue:43644|issue:43950", + "issue:43872|issue:44589", + "issue:46899|issue:46932", + "issue:43828|issue:44928", + "issue:43054|issue:45701", + "issue:45920|issue:46489", + "issue:38175|issue:43054", + "issue:45987|issue:46858", + "issue:43950|issue:45902", + "issue:43054|issue:43493", + "issue:44910|issue:45859", + "issue:45663|issue:45902", + "issue:42947|issue:44387", + "issue:44403|issue:44451", + "issue:24643|issue:28530", + "issue:43975|issue:45520", + "issue:44403|issue:44488", + "issue:43643|issue:45020", + "issue:41762|issue:42915", + "issue:41553|issue:43883", + "issue:42947|issue:43856", + "issue:44991|issue:45399", + "issue:43824|issue:43872", + "issue:45685|issue:45823", + "issue:45081|issue:45399" + ], + "prepared_review_unit_hash": "809141961f6eafe3736db428e0476a9f796bc0a817cb2f759a9573481d87a486", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36504, + "estimated_input_tokens": 9126, + "estimated_eval_tokens": 18508 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous issue cluster with a few true duplicates. The strongest duplicate family is the non-persistent-buffer regression around #43950/#44534, while #45902 is related thematically but not clearly the same underlying bug from the provided evidence.", + "confidence": 0.64, + "canonical_issue_reason": "#43950 is the clearest anchor for the non-persistent-buffer regression: it names the concrete transformers 5.x corruption bug directly and has an obvious near-duplicate in #44534.", + "canonical_pr_reason": null, + "best_issue_reason": "#43950 is the best issue to keep as the representative of the buffer-corruption subset because it is the most explicit and actionable description of that specific regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Different models and failure modes: tokenizer detokenization regression vs Qwen-3 loading error." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and do not read as the same concrete bug." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Docs/example errors with quantization versus a model-quality complaint; not the same underlying issue." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same regression family: non-persistent buffers being corrupted/junk-filled in transformers v5." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Different code paths and symptoms: custom compute_loss API breakage vs DeepSpeed ZeRO-3 loading failure." + }, + { + "left": "issue:43643", + "right": "issue:43950", + "accept": false, + "reason": "Remote-code config field loss is a different bug from non-persistent buffer corruption." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Distinct compatibility problems with different root causes and symptoms." + }, + { + "left": "issue:46899", + "right": "issue:46932", + "accept": true, + "reason": "Same Gemma4 quantized vision/audio embedder dtype-casting bug; #46932 is explicitly the fix/duplicate." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Different issues in different paths: autocast dtype mismatch versus 3D position_ids / SDPA fallback behavior." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "A model embedding quality complaint is not the same as a broad tokenizer-version change report." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Both involve tokenization, but they are not the same concrete tokenizer bug or model family." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Different SigLIP2 symptoms: unexpected zero probabilities versus worse text embeddings." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": true, + "reason": "Same StaticCache get_seq_length contract bug: tensor return value breaking downstream arithmetic." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "The titles are similar around buffers/materialization, but the packet does not clearly support the same underlying regression; this looks adjacent rather than duplicate." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Related model family, but one is a performance complaint and the other is an HF-vs-JAX discrepancy report." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Separate Qwen3.5 bugs: flash-attn crash from 3D position_ids versus CPU-offload incompatibility in rotary embedding." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Different model/runtime paths and distinct root causes; not enough evidence for a single bug." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Different performance/memory problems with different triggers and code paths." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic loading noise versus a specific model-loading failure." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "Unrelated training shape error versus early-stopping metric detection." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer detokenization bug versus import utility KeyError; unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Different loading reports: noisy logs versus a specific tokenizer/model load failure." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Specific missing-fields issue versus a broader remote-code regression claim." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Different model/runtime failures with different root causes." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad error messaging is not the same as a missing attribute error." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "LoRA gradient checkpointing ineffectiveness and Qwen3 MoE memory usage are different problems." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading regression versus flash-attn2 fallback selection logic; different subsystems." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Import/export symbol issue versus bitsandbytes constructor incompatibility." + }, + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "MPS histogram kernel issue is unrelated to Gemma4 device-map mismatch." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer regex patch crash versus fallback gating; no shared underlying bug." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster summary is mostly grounded, but one soft-edge acceptance looks overstated: issue:43950 vs issue:45902. The packet only shows title-level similarity around buffer/materialization/uninitialized memory; it does not clearly support calling them the same underlying regression. The other accepted pairs look conservative, but this one makes the verdict set non-conservative overall." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 5, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43782", + "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:44960", + "issue:44991", + "issue:45020", + "issue:45305", + "issue:45313", + "issue:45356", + "issue:45488", + "issue:45588", + "issue:45715", + "issue:45759", + "issue:45823", + "issue:45923", + "issue:46315", + "issue:46489", + "issue:46661", + "issue:46869", + "issue:46894", + "issue:46897" + ], + "soft_pairs": [ + "issue:43782|issue:44960", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:41553|issue:43881", + "issue:43828|issue:45823", + "issue:45020|issue:45356", + "issue:42491|issue:44368", + "issue:43881|issue:43883", + "issue:45715|issue:46315", + "issue:43872|issue:46869", + "issue:45759|issue:46661", + "issue:43872|issue:44291", + "issue:43827|issue:44509", + "issue:45488|issue:46489", + "issue:45020|issue:46661", + "issue:43404|issue:43950", + "issue:46894|issue:46897", + "issue:43792|issue:44220", + "issue:43257|issue:45305", + "issue:36331|issue:41093", + "issue:42947|issue:45923", + "issue:43940|issue:45313", + "issue:43854|issue:44863", + "issue:24643|issue:30064", + "issue:44479|issue:44560", + "issue:44928|issue:45588", + "issue:41093|issue:44805", + "issue:41553|issue:44991" + ], + "prepared_review_unit_hash": "3d69cc6b57e3a985e603a9a35e1cd84f903340493141142bb5b0f00694cb9756", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "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": "Mostly heterogeneous issue cluster with a few clear duplicate pairs. The Qwen video pair is related but not strongly evidenced as the same concrete bug, so it should be treated conservatively rather than merged as a duplicate.", + "confidence": 0.44, + "canonical_issue_reason": "Issue 45020 is the broadest representative of the cluster\u2019s dominant theme: recent-version regressions breaking model/tokenizer loading paths across multiple models. It is not a perfect umbrella, but it is the best single anchor for this mixed set.", + "canonical_pr_reason": null, + "best_issue_reason": "45020 is the strongest general representative because it captures a broad, centrally recurring regression pattern affecting loading behavior across models, which aligns better with the cluster than the narrower model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Different model families and failure modes: Qwen3-VL weight loading with `weight_only=True` vs GLM5 loading in tests." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Both are regressions, but one is remote-code/model loading and the other is a Gemma4 PLE forward device-mismatch bug." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Tokenizer mapping/add-new-model failure is a different code path from remote-code loading regressions." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different models and symptoms: Voxtral tokenizer error message vs GLM-4v-9b load failure." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Different runtime bugs in different models: autocast dtype mismatch vs device-map mismatch." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Remote-code loading regressions and tokenizer codec/warning regression are different issues." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA checkpoint compatibility on hf4.x/hf5.x is unrelated to the tie_word_embeddings warning." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Both are model-loading failures, but for different models and different missing/tied-weight problems." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "Both are tokenizer performance regressions, but they affect different conversion functions and are not clearly the same bug." + }, + { + "left": "issue:43872", + "right": "issue:46869", + "accept": false, + "reason": "Bitsandbytes `_is_hf_initialized` incompatibility is unrelated to YOLO multi-GPU device-map failures." + }, + { + "left": "issue:45759", + "right": "issue:46661", + "accept": false, + "reason": "Different loading bugs: missing `text_config` unwrap for composite configs vs remote-code `config_class` attribute error." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Both describe the same `_is_hf_initialized` compatibility break during model initialization/loading." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both are documentation cleanup for stale pipeline-task references removed in v5." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Both involve DeepSeek-family tokenizer regressions, but the specific tokenizer behavior and root cause differ." + }, + { + "left": "issue:45020", + "right": "issue:46661", + "accept": false, + "reason": "Remote-code model breakage and `config_class` AttributeError are related broadly, but not the same concrete bug." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Tied lm_head weight bug and non-persistent buffer corruption are distinct state/loading problems." + }, + { + "left": "issue:46894", + "right": "issue:46897", + "accept": false, + "reason": "Both are double-shift training-loss reports, but they are model-specific and not clearly the same concrete code-path bug." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Whisper loading failure and `_torch_extract_fbank_features()` issue are too indirect to treat as the same bug." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "DeepSpeed ZeRO-3 weight-loading failure and gradient-averaging bug are different training-time issues." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unexpected keyword in `compute_loss` and mask-shape IndexError are unrelated failures." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "One is PEFT LoRA gradient checkpointing not taking effect; the other is a missing `supports_gradient_checkpointing` flag." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are Qwen DeepSpeed ZeRO-3 load failures, but the affected model structures and missing-weights symptoms are different enough that the shared root cause is not clear." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures for unrelated model families." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "DeepSpeed weight-shape runtime error and void segmentation-map processing are unrelated bugs." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are Qwen/VL video regressions, but the packet only shows title-level similarity and does not establish the same concrete bug or a single mergeable fix." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "RLHF NaNs from position-id behavior and flash-attention `s_aux=None` crash are different problems." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": false, + "reason": "The error text is similar, but there is not enough evidence that these are the same underlying mask/indexing bug." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer-loading problems affecting different models and code paths." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Most of the verdicts are conservative, but one accepted soft edge is too speculative: issue 44479 and issue 44560 are both Qwen/VL video regressions, yet the packet only gives title-level evidence and does not clearly show they are the same concrete bug or a single mergeable PR. The summary also overstates this as a 'duplicate pair' rather than a related pair. Because the soft-edge acceptance is not fully grounded, I would reject." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 6, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:41553", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "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:46468", + "issue:46489", + "issue:46515", + "issue:46858", + "issue:46899" + ], + "soft_pairs": [ + "issue:43716|issue:46468", + "issue:43366|issue:43828", + "issue:44530|issue:44863", + "issue:44863|issue:46402", + "issue:43366|issue:43531", + "issue:43596|issue:46515", + "issue:45325|issue:45902", + "issue:43366|issue:44589", + "issue:41553|issue:44488", + "issue:43638|issue:46515", + "issue:44155|issue:45584", + "issue:45663|issue:45823", + "issue:42915|issue:43957", + "issue:43824|issue:43957", + "issue:43716|issue:44560", + "issue:43950|issue:45834", + "issue:42617|issue:43366", + "issue:43645|issue:43646", + "issue:44560|issue:45685", + "issue:44910|issue:46858", + "issue:44560|issue:44918", + "issue:36010|issue:42915", + "issue:44683|issue:46515", + "issue:43550|issue:44910", + "issue:45812|issue:46489", + "issue:45823|issue:46515", + "issue:44661|issue:45823", + "issue:44560|issue:44805", + "issue:38175|issue:43994", + "issue:44877|issue:45030", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:45672|issue:46899", + "issue:42491|issue:43720", + "issue:44661|issue:44805" + ], + "prepared_review_unit_hash": "557aeb2856dfa1a9c7f83fde8f2e57f38dd9aae618418bd2461c0c4646bc683c", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 38341, + "estimated_input_tokens": 9586, + "estimated_eval_tokens": 19428 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous, with only a few convincing duplicate pairs. The strongest matches are the Transformers 5.0 custom-model init regression, the SigLIP2 bad-output regression, the Deepspeed ZeRO-3 + BertModel IndexError, and the Gemma4 4-bit multimodal dtype-cast failure. Most other soft edges connect different models or different code paths and should stay separate.", + "confidence": 0.87, + "canonical_issue_reason": "issue:46899 is the clearest representative of a concrete, well-scoped regression: Gemma4 multimodal/audio processing breaks under 4-bit quantization due to unsafe dtype handling. It is more specific and actionable than the paired report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46899 is a strong global representative because it describes a precise user-facing failure, has a clear root-cause framing, and aligns with a valid duplicate candidate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43716", + "right": "issue:46468", + "accept": false, + "reason": "Different models and symptoms; one is a Mistral-3 image-preprocessor dtype mismatch, the other is a Nemotron example failure." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "Unrelated bugs: GGUF gpt-oss support vs a Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different failure modes and models: Qwen3.5 PagedAttentionCache crash vs NemotronH checkpoint loading." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Different model-loading issues in unrelated model families." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window handling are separate problems." + }, + { + "left": "issue:43596", + "right": "issue:46515", + "accept": false, + "reason": "No shared bug: ZeRO-3 BertModel IndexError vs a generic torch version incompatibility." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Both involve rotary-related code, but the models and root causes differ." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "Completely different topics: GGUF architecture support vs float8 storage lookup failure." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Different loading errors for different models; only broad tokenizer/loading subsystem overlap." + }, + { + "left": "issue:43638", + "right": "issue:46515", + "accept": false, + "reason": "ZeRO-3/BertModel IndexError is unrelated to a generic torch compatibility report." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Different audio bugs: track embedding leakage vs empty transcription handling." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Distinct Gemma4 issues: FSDP2 shared_kv_states rebuild vs device mismatch in PLE forward." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Different code paths: Qwen3Moe FP8 config failure vs meta-device loading issues." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Import error for Qwen2.5-VL is unrelated to meta-device loading regressions." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different models and errors; no shared underlying bug." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer loading regression and Kosmos2.5 OCR index error are unrelated." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "No concrete shared bug; these are different model/runtime issues." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same Transformers 5.0 custom-model initialization regression; the notebook-specific report is just a narrower reproduction." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Unrelated errors in different subsystems and model families." + }, + { + "left": "issue:44910", + "right": "issue:46858", + "accept": false, + "reason": "Both involve attention/cache internals, but the actual bugs and affected paths are different." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Different Qwen3.5-related failures with different triggers and code paths." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "Importing GenerationMixin and Qwen3Moe FP8 failures are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:46515", + "accept": false, + "reason": "Compiled flex_attention on torch >=2.9 is not the same bug as a torch 2.4.1 incompatibility report." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Both mention attention/compile, but the model, symptom, and likely root cause differ." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and likely different code paths." + }, + { + "left": "issue:45823", + "right": "issue:46515", + "accept": false, + "reason": "Gemma4 device mismatch is unrelated to a generic torch incompatibility issue." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "add-new-model-like TOKENIZER_MAPPING_NAMES failure and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different index/shape errors in different model pipelines." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2-base-patch16-224 bad-output regression; both report incorrect/degenerate outputs from the same model." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Strict config loading for granite_speech and glm4v config validation are separate regressions." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "ZeRO-3 BertModel IndexError is unrelated to tokenizer-mapping/model-addition logic." + }, + { + "left": "issue:45672", + "right": "issue:46899", + "accept": true, + "reason": "Both point to the same Gemma4 4-bit multimodal dtype-handling regression; the audio-only and audio/vision descriptions are variant reports of the same bug." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3_moe LoRA compatibility and BitNet accelerate-loading failure are different problems." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "No shared underlying bug: tokenizer mapping registration vs mask/index shape mismatch." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The accepted duplicate pairs are conservatively phrased and grounded in the packet, and the rejected soft edges are generally cautious rather than overstated. The summary is also supported by the issue titles, including the close Gemma4 and Transformers 5.0 matches. One note: the Deepspeed ZeRO-3/BertModel similarity is mentioned in the summary without a corresponding soft-edge verdict, but that does not appear to overstate the packet evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 7, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43425", + "issue:43632", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44756", + "issue:44863", + "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:45812", + "issue:45834", + "issue:46050", + "issue:46459", + "issue:46468", + "issue:46512", + "issue:46846", + "issue:46897", + "issue:46901" + ], + "soft_pairs": [ + "issue:43632|issue:45834", + "issue:43317|issue:44756", + "issue:38175|issue:39692", + "issue:43644|issue:44534", + "issue:41720|issue:42915", + "issue:42915|issue:44756", + "issue:42491|issue:45276", + "issue:44451|issue:44488", + "issue:44960|issue:44977", + "issue:44863|issue:46846", + "issue:45507|issue:46468", + "issue:45313|issue:45357", + "issue:46459|issue:46512", + "issue:43644|issue:43645", + "issue:44488|issue:44991", + "issue:39692|issue:43994", + "issue:43257|issue:43854", + "issue:43856|issue:43873", + "issue:42491|issue:45292", + "issue:42491|issue:45335", + "issue:43854|issue:43931", + "issue:43708|issue:44075", + "issue:43856|issue:44387", + "issue:44960|issue:46050", + "issue:44164|issue:44291", + "issue:45276|issue:45335", + "issue:46897|issue:46901", + "issue:45488|issue:45812", + "issue:45362|issue:45406", + "issue:43425|issue:43957" + ], + "prepared_review_unit_hash": "02ab5b0142f4a884e7e40445dd63f3c22f2cbae122b91652a4fd819f09b57fd9", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36464, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous Transformer/model-loading regressions. Only the non-persistent-buffer junk regression looks like a clear duplicate pair; the other soft pairs mostly share a subsystem or model family but not the same concrete bug.", + "confidence": 0.36, + "canonical_issue_reason": "issue:43644 is the clearest representative of the only strong duplicate pair: it states the broad Transformers v5 non-persistent-buffer corruption bug, while 44534 is the later rephrasing of the same regression.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the best issue candidate because it is the most general, clearly scoped duplicate target among the accepted pair and has the strongest standalone description.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "Unrelated bugs: _is_hf_initialized regression vs Kosmos2.5 OCR index error." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Different failure modes and code paths: device_map/offload loading vs mmap OOM mitigation." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Both involve SigLIP2, but one is zero probabilities and the other is doc/example mismatch plus quantization failure." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same concrete regression: Transformers v5 fills non-persistent buffers with junk; 44534 is the same bug restated." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Different issues: auto device-mapping CUDA assert vs FineGrainedFP8Config failure." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Unrelated Qwen3Moe FP8 config bug vs Strix Halo mmap/OOM workaround." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "LoRA hf4.x/hf5.x compatibility problem is unrelated to Gemma4 resize_token_embeddings behavior." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are load failures for different models, but no clear shared underlying bug from the titles alone." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and symptoms: GLM5 issue vs Qwen3.5 flash-attention generation bug." + }, + { + "left": "issue:44863", + "right": "issue:46846", + "accept": false, + "reason": "NemotronH checkpoint loading failure is not the same as Qwen3.5 TP plan OOM/shape errors." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "GraniteMoEHybrid invalid method call vs Nemotron example breakage are distinct problems." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5 regressions, but one is ZeRO-3 loading and the other is save_pretrained key serialization." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Different dtype issues: loading composite checkpoints vs from_config leaving config.dtype stale." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Non-persistent-buffer corruption is unrelated to Jupyter custom-model initialization." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer/model-loading failures for different model families; no evidence of the same bug." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Both are SigLIP2-related, but one is doc-example errors and the other is nonsensical inference outputs." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed is unrelated to GLM-4.7-Flash loading in tests." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Training memory inefficiency and quantization/offloading behavior are different bugs." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "The LoRA compatibility bug is unrelated to generic resize_token_embeddings not updating output embeddings." + }, + { + "left": "issue:42491", + "right": "issue:45335", + "accept": false, + "reason": "Different model/version-compatibility issue vs t5gemma embedding resize bug." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different models and symptoms: GLM-4.7-Flash load issue vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint-resume max_steps bug is unrelated to SGD optimizer args being ignored." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both mention memory, but one is training efficiency and the other is int4 quantization reserved-memory/OOM regression." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "GLM5 model issue and sinq save/load quantization regression are distinct." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "extra_state save/load handling is different from init_empty_weights/_is_hf_initialized loading TypeError." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Similar resize_token_embeddings theme, but different model families and affected submodules; not clearly the same concrete bug." + }, + { + "left": "issue:46897", + "right": "issue:46901", + "accept": false, + "reason": "Both are training-loss shift bugs, but they are model-specific and not proven to be the same code-path issue." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but the titles point to different mechanisms and model families, so they are not safely the same bug." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Qwen3.5 chat crash and Gemma4Processor _tokenizer AttributeError are different failures." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 compatibility and meta-device loading failures are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly duplicative pair in the titles is issue:43644 / issue:44534, which is an exact title match and supports the canonical choice. The other soft-edge rejections are appropriately cautious given the limited evidence in titles/bodies alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 8, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42947", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43454", + "issue:43645", + "issue:43854", + "issue:43873", + "issue:43957", + "issue:44079", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44488", + "issue:44589", + "issue:44756", + "issue:44863", + "issue:44877", + "issue:44910", + "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:44910|issue:45588", + "issue:44402|issue:44488", + "issue:39692|issue:44960", + "issue:44291|issue:44756", + "issue:41720|issue:44155", + "issue:45663|issue:45759", + "issue:43404|issue:43454", + "issue:42947|issue:45704", + "issue:43873|issue:44387", + "issue:44079|issue:46829", + "issue:45663|issue:45859", + "issue:44292|issue:44912", + "issue:42947|issue:43381", + "issue:36010|issue:39404", + "issue:43957|issue:44756", + "issue:44155|issue:45823", + "issue:43335|issue:45663", + "issue:44387|issue:44756", + "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:42175|issue:43957", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:44589|issue:44756", + "issue:44402|issue:44451" + ], + "prepared_review_unit_hash": "469a55762521db6027770b2ed7d879e72925a524e43883e7a369a1d06ad2cc0e", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37412, + "estimated_input_tokens": 9353, + "estimated_eval_tokens": 18962 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue set; the only clear duplicate cluster is the resize_token_embeddings propagation bug across related model variants.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 45292 is the cleanest canonical: it states the shared resize_token_embeddings/output_embeddings bug in the most general form, while the neighboring reports are model-specific variants.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the best representative because it is the broadest and most reusable description of the underlying bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both are Flash Attention bugs, but they fail in different code paths and for different causes." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading failures, but the reports name different models and do not show the same root cause." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "Different models and failure modes; one is a doc/example issue, the other is a GLM5 report." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "One is init_empty_weights argument handling; the other is mmap/OOM on Strix Halo." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Completely different subsystems and symptoms: device-map CUDA assert vs batched audio inference leakage." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Different model families and bugs: FSDP2 keyerror vs from_config text_config unwrapping." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both are weight-tying bugs, but for different model implementations and likely different fixes." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing/LoRA behavior and an Apex RMSNorm memory leak are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both mention quantization memory issues, but one is offloading and the other is int4 reserved-memory growth." + }, + { + "left": "issue:44079", + "right": "issue:46829", + "accept": false, + "reason": "ModelOutput key assignment and VideoPrism output formatting are different bugs." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Different model code paths: Gemma4 FSDP2 vs Qwen3_5Moe CPU offload incompatibility." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are quantization/loading problems, but they concern different models and quantization formats." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Same broad topic of training behavior, but one is checkpointing inefficacy and the other is eval-mode restriction." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin and Whisper pipeline return_language are unrelated." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading regressions and mmap/OOM on Strix Halo are different problems." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo batching leak and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and Gemma4 FSDP2 shared_kv_states are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Quantization OOM and mmap-on-Strix-Halo OOM have different causes and code paths." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint loading and the official Nemotron example break in different ways." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "Both report resize_token_embeddings failing to propagate new embeddings to output embeddings; same underlying bug." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video input regression in Qwen variants and strict config loading for granite_speech are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Eval-mode checkpointing restriction and Qwen3.5 RLHF gradient explosions are different issues." + }, + { + "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": "Both involve mistral_common tokenizers, but one is BPE marker handling and the other is continue_final_message logic." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Flash-attn sink-less crash and Gemma4 FSDP2 keyerror are different code paths." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "One is file permission propagation, the other is loading the wrong custom module after save_pretrained." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "GLM-4.7-Flash test loading and Qwen3.5 save_pretrained regression are different bugs." + }, + { + "left": "issue:42175", + "right": "issue:43957", + "accept": false, + "reason": "Package dependency/back-end installation and meta-device loading regressions are unrelated." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Both are resize_token_embeddings propagation failures in related model variants; same underlying class of bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Jupyter custom-model initialization and extra_state save/load handling are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Missing Float8 storage and Strix Halo mmap OOM are different failure modes." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both are tokenizer/model loading complaints, but not clearly the same root cause or code path." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted cluster, the resize_token_embeddings propagation bug across 45292/45276/45335, is well supported by the titles and looks like a single mergeable bugfix family. The other false verdicts do not overclaim shared root cause from the 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": 27, + "review_unit_index": 9, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:43299", + "issue:43425", + "issue:43452", + "issue:43596", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43716", + "issue:43828", + "issue:43856", + "issue:43901", + "issue:43906", + "issue:43931", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "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:46899", + "issue:46901" + ], + "soft_pairs": [ + "issue:43646|issue:43708", + "issue:46894|issue:46901", + "issue:43425|issue:43828", + "issue:46315|issue:46491", + "issue:45812|issue:46315", + "issue:44530|issue:45084", + "issue:39404|issue:45701", + "issue:43425|issue:46869", + "issue:43611|issue:43645", + "issue:44683|issue:45823", + "issue:43425|issue:43716", + "issue:39404|issue:46665", + "issue:43425|issue:45823", + "issue:44589|issue:45084", + "issue:44479|issue:44805", + "issue:43901|issue:44509", + "issue:43994|issue:44079", + "issue:43906|issue:44530", + "issue:44164|issue:45759", + "issue:36032|issue:43452", + "issue:44265|issue:44898", + "issue:46050|issue:46326", + "issue:43906|issue:46326", + "issue:44991|issue:45081", + "issue:44671|issue:44991", + "issue:39404|issue:44912", + "issue:45823|issue:46899", + "issue:45479|issue:46894", + "issue:44164|issue:46821", + "issue:43856|issue:44756", + "issue:43596|issue:44683", + "issue:43299|issue:43931" + ], + "prepared_review_unit_hash": "d4c5dbc9da89caf066374905557800f3e3dbacc8033d7af3b7c036011b851cd8", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37541, + "estimated_input_tokens": 9386, + "estimated_eval_tokens": 19028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is dominated by heterogeneous Transformers v5 regressions (tokenizers, model loading, multimodal dtype/device issues, and a few training/runtime bugs). None of the soft-edge pairs look like the same underlying defect closely enough to merge, so all proposed duplicate links should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "issue:45812 is the strongest canonical issue: it describes a concrete, broad v4\u2192v5 regression in tokenizer behavior with clear user impact across all Granite models, making it a better representative target than the more model-specific or more generic-loading issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the best overall issue in the set because it is concrete, broadly scoped, and clearly tied to a specific regression in a core code path (AutoTokenizer), which makes it a useful anchor for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Different failure modes: custom model init in notebooks vs Trainer resume/max_steps calculation. Same broad v5 breakage theme, but not the same bug." + }, + { + "left": "issue:46894", + "right": "issue:46901", + "accept": false, + "reason": "Both mention a double-shift pattern in training loss, but they affect different model families and no shared code-path is shown; too weak to treat as one bug." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "One is a Torch 2.10 compatibility complaint, the other is an autocast dtype mismatch in Phi-tiny-MoE. Different root causes and fixes." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Performance regression in convert_tokens_to_ids vs a round-trip tokenization correctness bug for CodeLlama. Not the same issue." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer IDs are wrong for Granite models, while this pair is about conversion performance for slow tokenizers. Different bug classes." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash vs a template-node compilation error. Both are runtime failures, but on different code paths and subsystems." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline return_language regression vs a general tokenization-change report. Too broad and not the same underlying defect." + }, + { + "left": "issue:43425", + "right": "issue:46869", + "accept": false, + "reason": "Torch version incompatibility and YOLO multi-GPU device_map failure are unrelated problems." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "base_model_prefix loading error and notebook custom-model initialization are both v5 model-loading issues, but they are not the same concrete bug." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch >=2.9 and Gemma4 device mismatch with device_map=\"auto\" are separate runtime issues." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch incompatibility vs Mistral-3 image preprocessor/model dtype mismatch. Different causes and code paths." + }, + { + "left": "issue:39404", + "right": "issue:46665", + "accept": false, + "reason": "Whisper pipeline regression and from_pretrained behavior inside Python classes are unrelated loading/pipeline issues." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "A Torch version incompatibility is not the same as a model forward device-placement bug." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup failure and template-node compilation failure are distinct low-level errors." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression for Qwen VL models vs a mask shape mismatch. Both are regressions, but not the same defect." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but one is about stale return_all_scores text and the other about removed pipeline task docs. Not the same issue." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Nonsensical SigLIP2 outputs and ModelOutput key assignment with None are different behavioral bugs." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Isolated reproduction of another issue vs a PagedAttentionCache crash. The first is just a reproduction report, not the same code-path bug." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "extra_state serialization failure and composite Qwen config unwrapping are unrelated save/load behaviors." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "T5 tokenizer add_special_tokens conflict and gguf_file failures in from_pretrained are both loading problems, but they arise from different mechanisms." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export failure due to torch_compilable_check vs Perceiver non-default-resolution interpolation failure. Different model families and failure modes." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantized sinq save/load corruption and mistral_common continue_final_message behavior are unrelated." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "A reproduction of another issue and a tokenizer prompt-construction bug are not the same underlying defect." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer loading failure for EMBEDDIA/est-roberta vs a Mistral regex patch crash. Both are tokenizer regressions, but on different code paths." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT prediction corruption and tokenizer load failure for est-roberta are distinct issues." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper return_language pipeline regression vs MXFP4 quantization fallback issue. Different models and logic." + }, + { + "left": "issue:45823", + "right": "issue:46899", + "accept": false, + "reason": "Gemma4 device mismatch during forward vs Gemma4 multimodal quantization dtype-cast bug. Related model family, but distinct failures and fixes." + }, + { + "left": "issue:45479", + "right": "issue:46894", + "accept": false, + "reason": "Zero-loss classification bug and ASR double-shift bug are different training-loss problems with no clear shared code path." + }, + { + "left": "issue:44164", + "right": "issue:46821", + "accept": false, + "reason": "extra_state save/load bug vs TimesFM window_size AttributeError. Different APIs and root causes." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE memory inefficiency and Strix Halo mmap OOM are both memory-related, but not the same defect." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Zero3/BertModel index error and compiled flex_attention torch>=2.9 failure are unrelated runtime problems." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3VL MoE loading breakage and Qwen3-VL-30B-A3B-Instruct weight-shape mismatch are similar only at a family level, but they are not the same concrete bug or 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": 27, + "review_unit_index": 10, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:32090", + "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:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44462", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45479", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:46178", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46489", + "issue:46897", + "issue:47050" + ], + "soft_pairs": [ + "issue:44488|issue:44987", + "issue:43632|issue:47050", + "issue:45479|issue:46897", + "issue:44488|issue:45081", + "issue:43611|issue:43644", + "issue:36032|issue:42915", + "issue:44462|issue:46489", + "issue:42796|issue:43404", + "issue:43716|issue:45084", + "issue:46285|issue:46402", + "issue:44661|issue:44683", + "issue:32090|issue:39404", + "issue:44451|issue:44987", + "issue:43643|issue:44530", + "issue:43541|issue:45084", + "issue:42915|issue:44530", + "issue:44737|issue:45859", + "issue:40990|issue:43421", + "issue:34567|issue:40444", + "issue:43644|issue:43646", + "issue:42175|issue:46178", + "issue:40444|issue:46326", + "issue:36032|issue:43645", + "issue:43582|issue:43632", + "issue:44206|issue:44220", + "issue:44987|issue:44991", + "issue:43452|issue:43482", + "issue:44811|issue:45584", + "issue:45084|issue:45507", + "issue:43317|issue:43482" + ], + "prepared_review_unit_hash": "b665da7647bb5d814690c8d4a5616f4407550c9e0bb12e1978cb6ca6d2804fe3", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36791, + "estimated_input_tokens": 9198, + "estimated_eval_tokens": 18652 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct regressions across tokenizer loading, model initialization, audio feature extraction, and compile/runtime paths. I found two plausible duplicate pairs that appear to describe the same concrete loader/helper bug rather than just the same subsystem.", + "confidence": 0.67, + "canonical_issue_reason": "Issue 43644 is a clear, self-contained Transformers v5 regression with a concrete symptom and broad impact, making it a reasonable anchor issue.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43644 is the strongest representative issue here: it describes a specific v5 regression, is easy to understand, and has some discussion/reference activity.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both are tokenizer/model loading failures, but they affect different repositories and likely different root causes." + }, + { + "left": "issue:43632", + "right": "issue:47050", + "accept": false, + "reason": "These are unrelated bugs: an initialization flag regression versus an ONNX attention dtype mismatch." + }, + { + "left": "issue:45479", + "right": "issue:46897", + "accept": false, + "reason": "Both mention training loss problems, but they involve different models and the titles do not indicate the same code path." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer-loading bugs: one is a repo loading failure, the other is a Mistral regex patch crash." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Both are v5 regressions, but one is about base_model_prefix loading and the other about non-persistent buffer initialization." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "These are unrelated tokenizer and FP8-config failures." + }, + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "Both concern tokenizers, but one is about ignoring tokenizer.json while the other is an output regression in a specific model tokenizer." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Different Mistral-related bugs: tokenizer BPE markers versus tied lm_head weights in a conditional generation model." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "These are different failures with no clear shared code path: image preprocessor dtype mismatch versus template compilation errors." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "Different model-structure/weight-loading regressions." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "An add-new-model registry bug and a flex_attention compile regression are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast and Whisper return_language pipeline issues are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Both are loading failures, but they appear to be different models and different tokenizer/model paths." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "A trust_remote_code AutoConfig regression is unrelated to the Qwen3.5 PagedAttentionCache crash." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Both involve compilation/tracing, but the concrete failures and code paths are different." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Different Qwen-family bugs affecting different features and models." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "Both are device/offload-related, but the model-specific bugs are distinct." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity/calibration issues and runtime post-processor updates are not the same bug." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "Unrelated trainer token-count tracking and iterable-dataset multi-image failure." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 initialization regressions, but the titles point to different symptoms and no shared concrete code path." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "These are different backend-mapping/package-installation issues." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "A multimodal training failure and a mistral_common tokenizer message-handling bug are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "T5 tokenizer method conflicts and custom-model initialization in notebooks are different problems." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "An Apple Silicon allocator TypeError is unrelated to the `_is_hf_initialized` regression." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": true, + "reason": "These look like the same audio feature-extraction regression: one names the unsupported center arg, the other points to the internal `_torch_extract_fbank_features()` helper that would surface it." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are v5 tokenizer-loading failures, but they involve different models and no shared concrete failing path is evident." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": true, + "reason": "Both are GGUF loading regressions under Transformers v5; 43482 looks like a specific model manifestation of the broader gguf_file/from_pretrained breakage in 43452." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both involve Whisper special-token behavior, but one is batch_decode skipping tokens and the other is empty-transcription generation after align_special_tokens." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "A template compilation error and a GraniteMoEHybrid invalid method call are different bugs." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "GPU/CPU offload loading failure and GGUF loading failure are different regression types." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The two accepted soft pairs are plausible duplicates based on the titles and share a concrete bug theme without overclaiming a shared code path. I don\u2019t see any clear evidence inflation that would justify rejecting." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 11, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28530", + "issue:34567", + "issue:36032", + "issue:41720", + "issue:42915", + "issue:43208", + "issue:43262", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43540", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43746", + "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:44488", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44849", + "issue:44877", + "issue:44987", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45987", + "issue:46326", + "issue:46661", + "issue:46821", + "issue:46869", + "issue:46899", + "issue:47013" + ], + "soft_pairs": [ + "issue:43746|issue:46326", + "issue:43262|issue:43377", + "issue:43792|issue:46869", + "issue:43208|issue:47013", + "issue:44403|issue:45081", + "issue:42915|issue:45084", + "issue:43645|issue:46661", + "issue:43646|issue:44534", + "issue:43377|issue:43819", + "issue:44849|issue:45325", + "issue:44987|issue:45081", + "issue:28530|issue:34567", + "issue:42915|issue:44291", + "issue:43404|issue:44466", + "issue:36032|issue:43742", + "issue:44451|issue:45081", + "issue:43883|issue:44488", + "issue:43994|issue:44610", + "issue:44062|issue:44987", + "issue:44464|issue:45987", + "issue:42915|issue:44589", + "issue:44008|issue:46899", + "issue:44265|issue:44610", + "issue:43531|issue:44877", + "issue:36032|issue:41720", + "issue:43632|issue:46821", + "issue:43322|issue:43540", + "issue:44479|issue:45325", + "issue:43425|issue:44403" + ], + "prepared_review_unit_hash": "81ce28a1efea6e47d2d9ed4e2b322abd696c1f8a707a0620d19b3b8f8c03c617", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36564, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 + }, + "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 with superficial text similarity (mostly Transformers v5, multimodal models, tokenizer loading, and xLSTM). None of the soft pairs look like the same underlying bug/change, so no duplicates should be merged.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue: the items span distinct bugs in different code paths and model families, so collapsing them would be misleading.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a good global representative of this cluster; the similarity is mostly topical rather than duplicate-level.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43746", + "right": "issue:46326", + "accept": false, + "reason": "Both involve model loading/tokenization workflows, but one is PEFT local checkpoint loading for GraniteSpeech while the other is mistral_common continue_final_message behavior; different bugs." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio processor chat-template sampling-rate defaulting and MIMI batched-vs-single output mismatch are unrelated failures in different components." + }, + { + "left": "issue:43792", + "right": "issue:46869", + "accept": false, + "reason": "Whisper load failure and YOLO multi-GPU device-map failure are separate model/runtime issues with different symptoms and code paths." + }, + { + "left": "issue:43208", + "right": "issue:47013", + "accept": false, + "reason": "Both are xLSTM, but one reports multiple training bugs for small models and the other is a specific return_last_states unpacking crash; not the same change." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Generic loading noise and a mistral regex patch crash are different problems; one is user-facing logging, the other tokenizer patch logic." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8Config failure and non-template-node compilation error are distinct multimodal/compiler issues." + }, + { + "left": "issue:43645", + "right": "issue:46661", + "accept": false, + "reason": "Both concern custom/remote-code model initialization, but one is notebook/custom model init with v5, while the other is missing config_class on remote-code models; different root causes." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Custom model initialization breakage and non-persistent buffer junk are separate Transformer v5 regressions." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "MIMI padding-mask batching mismatch is unrelated to DAC.from_latents vs forward/STE consistency." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and Qwen2.5-VL rope-index scaling bug affect different models and behaviors." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Loading physical-intelligence/fast failing on v5.1.0 is unrelated to the mistral tokenizer regex patch crash." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Early stopping/eval metric handling and TrainerState token counting are different Trainer bugs with no shared code path indicated." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Both mention v5 / initialization, but FineGrainedFP8Config Qwen3Moe failure is unrelated to init_empty_weights and _is_hf_initialized argument handling." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Mistral3 lm_head weight tying and device-dependent serialization of lm_head.weight are related only at a broad 'tied weights' level, but they describe different bugs and fixes." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "T5 tokenizer add_special_tokens conflict and MobileLLM key error are separate tokenizer/model-loading issues." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "ScandiBERT loading failure and mistral regex patch crash involve different models and failure modes." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Molmo missing all_tied_weights_keys and sleng-bert loading failure are unrelated model-loading regressions." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Siglip2 nonsensical outputs and OmDet-Turbo processor size mismatch are different vision-model bugs." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken special-argument duplication and physical-intelligence/fast loading failure are unrelated tokenizer vs model-loading issues." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Chunked generation inconsistency under compiled forward and StaticCache.get_seq_length contract violation are different generation/cache bugs." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Float8 storage lookup failure are distinct quantization/runtime issues." + }, + { + "left": "issue:44008", + "right": "issue:46899", + "accept": false, + "reason": "Gemma 3n audio/vision variable collision and Gemma4 multimodal 4-bit dtype cast issue are different multimodal bugs." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export with torch_compilable_check and OmDet-Turbo processor resolution mismatch are unrelated export vs preprocessing problems." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug and granite_speech strict config loading are separate model/config issues." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "T5 tokenizer add_special_tokens conflict and Qwen3 auto device mapping cudaErrorAssert are unrelated." + }, + { + "left": "issue:43632", + "right": "issue:46821", + "accept": false, + "reason": "_is_hf_initialized flag breakage and TimesFM window_size AttributeError are different Transformers v5 regressions in separate subsystems." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Llava Next segmentation fault and Qwen3OmniMoe video-input ValueError are distinct multimodal loading/processing bugs." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Qwen2.5/VL video-input regression and Qwen2.5-VL rope-index temporal scaling bug are both VL-related but affect different code paths and symptoms." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch 2.10 incompatibility is a dependency-version issue; unnecessary loading noise is a logging/polish 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": 27, + "review_unit_index": 12, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43089", + "issue:43296", + "issue:43381", + "issue:43421", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43716", + "issue:43746", + "issue:43792", + "issue:43883", + "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:45216", + "issue:45325", + "issue:45357", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:45987", + "issue:46315", + "issue:46491", + "issue:46710", + "issue:46821", + "issue:46899" + ], + "soft_pairs": [ + "issue:43089|issue:45987", + "issue:43296|issue:43531", + "issue:45216|issue:45357", + "issue:40444|issue:45325", + "issue:43611|issue:44534", + "issue:44220|issue:44683", + "issue:44991|issue:46315", + "issue:43716|issue:46899", + "issue:44403|issue:44991", + "issue:43883|issue:44991", + "issue:43381|issue:45923", + "issue:42796|issue:43746", + "issue:45812|issue:46491", + "issue:43792|issue:44589", + "issue:44291|issue:44589", + "issue:43632|issue:44756", + "issue:33666|issue:40444", + "issue:40990|issue:44811", + "issue:42796|issue:46710", + "issue:43482|issue:43525", + "issue:35707|issue:40444", + "issue:44079|issue:45663", + "issue:43531|issue:44291", + "issue:43421|issue:44568", + "issue:41720|issue:45081", + "issue:42175|issue:45542", + "issue:43482|issue:45538", + "issue:44462|issue:45812", + "issue:43296|issue:46821", + "issue:44079|issue:44534", + "issue:42175|issue:43296" + ], + "prepared_review_unit_hash": "1e057b8b5e33ac06e8fcbc0c981d7f85423903e44290fdbe60123e3e617789de", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36498, + "estimated_input_tokens": 9125, + "estimated_eval_tokens": 18506 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous rather than one tight duplicate cluster. Only a couple of pairs look like true duplicates: the Qwen3.5 save_pretrained regression pair and the TensorFlow-backend packaging pair.", + "confidence": 0.76, + "canonical_issue_reason": "No single canonical issue fits the whole cluster; the items span unrelated tokenizer, generation, multimodal, packaging, and model-loading bugs. If forced to pick one umbrella issue for the only clear subcluster, issue 45216 is the best representative because it is the broader Qwen3.5 save_pretrained regression and 45357 looks like a narrower follow-up.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45216 is the strongest issue candidate: it is a clear, user-facing regression report with a broader scope and it plausibly subsumes 45357. The rest of the items are too mixed to share one canonical issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Different problems: generation runtime overhead vs a StaticCache return-type contract bug." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "Different model families and failure modes: PaddleOCR-VL loading in vLLM vs Qwen3-MoE sliding_window behavior." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Both report the same Qwen3.5 save_pretrained regression; 45357 looks like a narrower symptom of incorrect checkpoint contents." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen2.5-VL-related, but one is multi-image training failure and the other is rope/position_id scaling for still images." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Both involve Transformers v5 regressions, but they affect different code paths: model loading with base_model_prefix vs non-persistent buffer initialization." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Unrelated subsystems: fbank feature extraction vs compiled flex_attention on newer torch." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Both touch tokenizers in v5, but one is tokenizer loading correctness and the other is slow-tokenizer performance with many added tokens." + }, + { + "left": "issue:43716", + "right": "issue:46899", + "accept": false, + "reason": "Both are multimodal dtype issues, but they involve different models and different failure paths (preprocessor/model mismatch vs unsafe cast in 4-bit embedders)." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Different bugs: noisy model loading vs tokenizer loading failure for a specific repo." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Different model-loading attribute error vs tokenizer-loading regression." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "One is about checkpointing being disallowed in eval mode; the other is a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Different issues: tokenizer BPE marker handling vs PEFT/local-checkpoint loading for GraniteSpeech." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different behavior (wrong token IDs vs encode/decode space stripping)." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Different failures: Whisper model execution vs missing Float8 storage registration." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Different bugs: init_empty_weights/_is_hf_initialized argument mismatch vs Float8 storage lookup failure." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Unrelated: _is_hf_initialized flag regression vs mmap/OOM on Strix Halo." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both Qwen2-VL/Qwen2.5-VL training-related, but multi-GPU training and multi-image prompt failure are separate problems." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Different models and symptoms: high perplexity on gpt-oss vs Whisper batch_decode skipping special tokens." + }, + { + "left": "issue:42796", + "right": "issue:46710", + "accept": false, + "reason": "Both are tokenizer/model quality regressions, but one is a tokenizer marker bug and the other is an accuracy regression after version upgrade." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Unrelated model-loading failures: GGUF load failure vs missing pad_token_id on Llama4Config." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Different generation mechanics: progressive generation with inputs_embeds/past_key_values vs finetuning with multiple images per prompt." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "Different bugs: ModelOutput key assignment vs Gemma4 FSDP2 shared_kv_states rebuild issue." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Different code paths and errors: Qwen3-MoE sliding_window vs init_empty_weights/_is_hf_initialized mismatch." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both involve special tokens, but one is runtime post-processor updates and the other is add_special_tokens behavior for a specific tokenizer." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Different failures: Qwen3 device mapping cuda assert vs Mistral regex patch crashing on tokenizer object shape." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": true, + "reason": "Both describe the same packaging problem: installing the torch extra leaves the TensorFlow backend unavailable/half-installed." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Unrelated: GGUF loading regression vs CLIPTokenizer model_max_length behavior." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both are AutoTokenizer regressions, but one ignores repository tokenizer.json while the other returns wrong token IDs for Granite models." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "Different model families and failures: PaddleOCR-VL/vLLM loading vs TimesFM window_size AttributeError." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "Different regression classes: ModelOutput key handling vs non-persistent buffer initialization in v5." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Different domains: backend packaging for TensorFlow vs loading a specific multimodal model in vLLM." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The cluster is indeed heterogeneous, and the two accepted pairs look like plausible same-bug duplicates: 45216/45357 are both Qwen3.5 save_pretrained regressions, and 42175/45542 are both TensorFlow-backend packaging issues. The canonical-issue choice is cautious and supported by the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 13, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42796", + "issue:42915", + "issue:43262", + "issue:43317", + "issue:43377", + "issue:43454", + "issue:43482", + "issue:43493", + "issue:43582", + "issue:43606", + "issue:43611", + "issue:43644", + "issue:43646", + "issue:43701", + "issue:43746", + "issue:43818", + "issue:43828", + "issue:43856", + "issue:43873", + "issue:43994", + "issue:44079", + "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:45491", + "issue:45584", + "issue:45701", + "issue:45823", + "issue:45834", + "issue:46402", + "issue:46821", + "issue:46909", + "issue:47217" + ], + "soft_pairs": [ + "issue:44509|issue:46909", + "issue:43317|issue:43856", + "issue:43611|issue:43646", + "issue:43377|issue:45491", + "issue:43644|issue:44079", + "issue:43818|issue:46402", + "issue:44265|issue:45072", + "issue:40444|issue:45381", + "issue:43818|issue:45357", + "issue:43493|issue:43994", + "issue:45325|issue:45834", + "issue:45325|issue:45381", + "issue:44829|issue:45479", + "issue:43454|issue:44466", + "issue:40444|issue:47217", + "issue:44291|issue:44877", + "issue:42915|issue:43482", + "issue:43701|issue:45325", + "issue:43873|issue:45701", + "issue:42796|issue:45381", + "issue:43606|issue:44265", + "issue:40990|issue:45584", + "issue:44479|issue:45823", + "issue:43828|issue:46821", + "issue:43262|issue:43746", + "issue:35707|issue:41720", + "issue:41553|issue:43582", + "issue:34634|issue:35707" + ], + "prepared_review_unit_hash": "7559fcd6ca482a0ef65d9018297864b629d9a4540061571512b83d249853492a", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36258, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of unrelated Transformers/model bugs, not a tight duplicate set. The only recurring pattern is broad v5 loading/init regressions, but the individual reports usually hit different code paths and models.", + "confidence": 0.37, + "canonical_issue_reason": "No true duplicate canonical exists; the closest broad anchor is issue 43611 because it represents the recurring v5 model-loading/init regression theme.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43611 is the broadest and most central representative of the set, so it is the best issue anchor even though it does not subsume the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44509", + "right": "issue:46909", + "accept": false, + "reason": "Both are documentation problems, but they concern different docs and different broken examples." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "One is a loading/offload failure; the other is a training memory-usage report for a different model." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both involve v5 model initialization/loading, but they are distinct failures with different triggers and fixes." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both are batching/padding-related, but they affect different models and different failure modes." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "One is about non-persistent buffers in serialization; the other is ModelOutput key assignment logic." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Different models and different bugs: missing video_tower temporal attention vs text encoder weights not loading." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "Export failure from torch_compilable_check is unrelated to the bf16 dtype-mismatch inference bug." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL multimodal issues, but one is iterable-dataset multi-image finetuning and the other is video position-id computation." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Same broad family, but one is a Video-LLaVA architecture issue and the other is Qwen3.5 save_pretrained key handling." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": false, + "reason": "Both mention SigLIP2, but one is an implementation discrepancy and the other is a nonsensical-output report; not clearly the same bug." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Both are index/position problems, but in different models and different input paths." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Both touch Qwen2.5-VL position handling, but one is still-image RoPE scaling and the other is video vision_position_ids." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both are training/loss regressions, but one is flash-attention-3-specific and the other is a single-label classification edge case." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both involve lm_head/tied weights, but one is a model config/load bug and the other is a serialization inconsistency." + }, + { + "left": "issue:40444", + "right": "issue:47217", + "accept": false, + "reason": "Both are VL input handling issues, but one is dataset/multi-image finetuning and the other is image-placeholder-in-text parsing." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "Both are v5 loading regressions, but they stem from different strictness/argument-handling paths." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Different quantization/loading failures: FineGrainedFP8Config vs GGUF loading under v5." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen2.5-VL RoPE/position-id handling." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to a tokenization-version regression." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL video position IDs." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU-offload device mismatch and torch.export/compilability are different code paths." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "Perplexity regression on gpt-oss-20b is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Both are inference regressions, but one is Qwen video-input handling and the other is Gemma4 device-map forward placement." + }, + { + "left": "issue:43828", + "right": "issue:46821", + "accept": false, + "reason": "An autocast dtype mismatch is unrelated to the TimesFM window_size AttributeError." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio processor chat-template sample-rate defaults and GraniteSpeech PEFT checkpoint loading are different issues." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to the Qwen3 auto-device-map CUDA assert." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Bad AutoTokenizer error messaging and an AppleSilicon allocator TypeError are unrelated." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are different 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": 27, + "review_unit_index": 14, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:43550", + "issue:43572", + "issue:43606", + "issue:43720", + "issue:43749", + "issue:43819", + "issue:43873", + "issue:44186", + "issue:44265", + "issue:44387", + "issue:44466", + "issue:44534", + "issue:44568", + "issue:44610", + "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:44186|issue:44265", + "issue:41720|issue:45823", + "issue:43606|issue:43819", + "issue:43454|issue:43550", + "issue:43257|issue:45923", + "issue:45823|issue:46821", + "issue:42796|issue:46209", + "issue:43377|issue:44610", + "issue:44568|issue:45920", + "issue:43454|issue:45072", + "issue:42617|issue:42915", + "issue:40990|issue:42947", + "issue:46305|issue:46315", + "issue:43377|issue:44898", + "issue:43454|issue:43720", + "issue:43749|issue:44534", + "issue:43572|issue:45759", + "issue:43317|issue:43873", + "issue:36331|issue:42617", + "issue:41720|issue:43322", + "issue:43322|issue:46050", + "issue:44877|issue:44991", + "issue:36331|issue:42915", + "issue:41720|issue:44977", + "issue:44387|issue:45704", + "issue:40990|issue:45701", + "issue:41720|issue:42617", + "issue:43450|issue:45701", + "issue:43065|issue:46402", + "issue:44466|issue:45357", + "issue:26413|issue:28530" + ], + "prepared_review_unit_hash": "3c0672571603c07df8342419f94e6fa0b05ebe1f6a410e7a45b250390454f6fc", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37063, + "estimated_input_tokens": 9266, + "estimated_eval_tokens": 18788 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated closed issues that only share broad subsystem similarity (loading, tokenization, offloading, quantization, training-loop regressions). I found no true duplicate pairs among the soft candidates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 43317 is the clearest representative of the shared theme around model loading/offload regressions, with a concrete and actionable failure mode; however, the cluster as a whole is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43317 is the strongest standalone issue here because it describes a specific, user-relevant regression in a common code path (`device_map=auto` + offload) without being overly model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Both are export/processing failures, but one is a tokenizer crash on NER batching while the other is `torch.export` failing on `torch_compilable_check`; different code paths and fixes." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Both mention auto device mapping, but the concrete bugs differ: CUDA assert on Qwen3 vs a Gemma4 PLE device mismatch during forward." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "One is a CPU offload device-mismatch bug for `suno/bark-small`; the other is a DAC latent decoding mismatch caused by missing STE. Same broad area, different defect." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "AyaVision lm_head tying failure and Bamba torch.compile+SDPA failure are unrelated model/runtime bugs." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Qwen3 MoE loading with accelerate+deepspeed and a missing `supports_gradient_checkpointing` flag in Nemotron are separate issues." + }, + { + "left": "issue:45823", + "right": "issue:46821", + "accept": false, + "reason": "Gemma4 device mismatch under `device_map=\"auto\"` and TimesFM `window_size` AttributeError are different models and different failure modes." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "MIMI batch-vs-single padding-mask inconsistency and OmDet-Turbo processor output-size mismatch affect different components and bugs." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "One is a single tokenizer missing BOS/EOS insertion; the other is a broader v5 regression producing wrong token IDs across many tokenizers." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "AyaVision lm_head tying/garbage text and dtype mismatches in SwitchTransformers/TimmWrapperModel are unrelated." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "A `3d_parallel.py` runtime failure is not the same as Qwen3Moe failing with `FineGrainedFP8Config`." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "High perplexity on gpt-oss and ineffective gradient checkpointing with PEFT LoRA are different symptoms and code paths." + }, + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "MoonshineStreaming mask off-by-one and tokenizer conversion slowdown are unrelated regressions." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "MIMI batching/padding-mask behavior and Perceiver interpolation on non-default resolution are distinct model bugs." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "lm_head tying issues are unrelated to 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 both v5-loading related, but they are not the same concrete bug." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing `pad_token_idx` in a config and failure to unwrap `text_config` in composite Qwen configs are different config-loading problems." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both involve offloading/loading, but 43317 is a specific `device_map=auto` dequantized-model failure while 43873 is a broader quantization/offloading behavior report." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "`compute_loss()` kwarg incompatibility and a `3d_parallel.py` failure are unrelated training/script issues." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 auto device map assert and Llava Next segmentation fault are different models and failure mechanisms." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "A load-time segfault for Llava Next and broken quantized save/load after `sinq` are different bugs." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config blocking `granite_speech` and tokenizer loading for `est-roberta` are separate config/tokenizer issues." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "The `compute_loss` signature mismatch is unrelated to Qwen3Moe `FineGrainedFP8Config` failures." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Auto device-map CUDA assert in Qwen3 and flash-attention generation issues in Qwen3.5 are distinct runtime bugs." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "CUDA reserved-memory OOM under int4 quantization and an apex RMSNorm memory leak are not the same problem." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "High perplexity on a specific model and version-dependent tokenization changes are different regressions." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "A model-loading CUDA assert is not the same as a `3d_parallel.py` execution failure." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape errors and tokenizer-version behavior changes are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and missing Sam3 text-encoder weights are different issues in different parts of the model." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "lm_head serialization inconsistency and incorrect visual-encoder keys on save_pretrained are distinct save/load regressions." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Checkpoint-resume epoch iterator failure and early-stopping metric detection are different trainer 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": 27, + "review_unit_index": 15, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28530", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43317", + "issue:43404", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43646", + "issue:43708", + "issue:43720", + "issue:43873", + "issue:43881", + "issue:44008", + "issue:44186", + "issue:44423", + "issue:44451", + "issue:44464", + "issue:44488", + "issue:44610", + "issue:44671", + "issue:44734", + "issue:44743", + "issue:44756", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45685", + "issue:45701", + "issue:45823", + "issue:46291", + "issue:46612", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:46869", + "issue:47020" + ], + "soft_pairs": [ + "issue:43881|issue:44488", + "issue:43873|issue:44756", + "issue:44186|issue:45072", + "issue:46612|issue:46858", + "issue:45685|issue:46821", + "issue:43295|issue:45701", + "issue:43550|issue:43720", + "issue:43577|issue:43708", + "issue:36010|issue:42617", + "issue:43646|issue:46665", + "issue:44451|issue:47020", + "issue:43550|issue:43606", + "issue:44186|issue:44610", + "issue:42915|issue:43317", + "issue:43404|issue:45357", + "issue:46821|issue:46869", + "issue:44610|issue:44898", + "issue:28530|issue:33666", + "issue:41720|issue:45161", + "issue:44734|issue:46858", + "issue:43531|issue:43572", + "issue:44008|issue:45823", + "issue:44423|issue:45406", + "issue:42915|issue:46291", + "issue:36331|issue:41720", + "issue:44464|issue:46858", + "issue:43881|issue:44991", + "issue:44671|issue:44743" + ], + "prepared_review_unit_hash": "bc1e1547e04ae46f7fb9f6491d5cb078f3772ea06c0e1d2ae32575bd9bc9eda7", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35575, + "estimated_input_tokens": 8894, + "estimated_eval_tokens": 18044 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a loose similarity cluster around Transformers regressions in loading, generation, cache, and quantization, but the soft pairs do not look like true duplicates of the same underlying bug. No PRs are present.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43646 is the broadest representative of the cluster: a Transformers 5.0 regression breaking custom model initialization and related downstream loading/custom-code failures.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best standalone exemplar because it captures the widest recurring theme in the cluster (breaking API/init behavior after version changes) without being overly model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Both are model loading failures, but they involve different models and likely different root causes." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both touch offload/memory behavior, but one is quantization offloading and the other is mmap/OOM on a specific platform; not the same bug." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer/NER padding crash vs bfloat16 dtype mismatch are different failure modes." + }, + { + "left": "issue:46612", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache internals, but one is beam-search reorder and the other is seq-length tensor arithmetic in static cache; different code paths." + }, + { + "left": "issue:45685", + "right": "issue:46821", + "accept": false, + "reason": "MPS histogram/int error and TimesFM window_size AttributeError are unrelated bugs." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Both are version-related regressions, but they concern different API breakages and behaviors." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile/SDPA failure and packed-weight loading failure are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Model dtype loading bug vs resume_from_checkpoint max_steps miscalculation are different issues." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "ImportError for GenerationMixin and inability to run a script are not the same underlying bug." + }, + { + "left": "issue:43646", + "right": "issue:46665", + "accept": false, + "reason": "Both mention custom model/from_pretrained behavior, but the reported failures and code paths are too different to merge." + }, + { + "left": "issue:44451", + "right": "issue:47020", + "accept": false, + "reason": "Different tokenizer-loading failures on different models; no concrete shared root cause is established." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "torch.compile with SDPA and CPU offload device mismatch are separate runtime bugs." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Tokenizer crash and image-size mismatch are both preprocessing-related, but not the same defect." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "FP8 fine-grained config failure and dequantized model offload failure are distinct quantization/loading problems." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "lm_head tying bug and save_pretrained visual-key regression are different state/serialization issues." + }, + { + "left": "issue:46821", + "right": "issue:46869", + "accept": false, + "reason": "TimesFM argument handling and YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Both concern image resolution handling, but they affect different models and different preprocessing/forward logic." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early stopping metric configuration and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both are parallelism/device-map problems, but one is a cuda assert and the other is tensor parallelism support; not the same bug." + }, + { + "left": "issue:44734", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache indexing, but one is serve KV continuation and the other is static-cache prefill arithmetic on non-CUDA backends." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and StableLmConfig pad_token_idx regression are unrelated." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Gemma3n forward variable collision and Gemma4 PLE device mismatch are different forward-pass bugs." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both hit transformers serve with multimodal models, but the crash causes and failing paths are different." + }, + { + "left": "issue:42915", + "right": "issue:46291", + "accept": false, + "reason": "FP8 model failure and kernels package import/version requirement are unrelated." + }, + { + "left": "issue:46612", + "right": "issue:46858", + "accept": false, + "reason": "Same general cache area, but distinct defects and different observable failures." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Trainer compute_loss signature change and multi-GPU cuda assert are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:46858", + "accept": false, + "reason": "Both mention generation/cache and compiled forward, but the reported bugs are not the same concrete issue." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer/model loading failures with no shared root cause." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "Masked LM prediction regression and recurrent-state reset in cache are different 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": 27, + "review_unit_index": 16, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:43425", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43873", + "issue:43883", + "issue:44186", + "issue:44361", + "issue:44492", + "issue:44509", + "issue:44534", + "issue:44589", + "issue:44913", + "issue:45059", + "issue:45406", + "issue:45464", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:46665", + "issue:46821", + "issue:47020", + "issue:47145" + ], + "soft_pairs": [ + "issue:43526|issue:43883", + "issue:45632|issue:45684", + "issue:44186|issue:44361", + "issue:43425|issue:46821", + "issue:28530|issue:30064", + "issue:43316|issue:44913", + "issue:45701|issue:47145", + "issue:42915|issue:43322", + "issue:33357|issue:35707", + "issue:29159|issue:36032", + "issue:30064|issue:34634", + "issue:43873|issue:45704", + "issue:43540|issue:43782", + "issue:43531|issue:45759", + "issue:44492|issue:44509", + "issue:43550|issue:43653", + "issue:43452|issue:47020", + "issue:44534|issue:46665", + "issue:41720|issue:43746", + "issue:43606|issue:43720", + "issue:42490|issue:43329", + "issue:34567|issue:35707", + "issue:43531|issue:46821", + "issue:34634|issue:41720", + "issue:45406|issue:45464", + "issue:43606|issue:43653", + "issue:43388|issue:45059", + "issue:43531|issue:44589", + "issue:45759|issue:45859" + ], + "prepared_review_unit_hash": "331842db69c58689eba96bc0fa74a6e9f949ecf4b4b73a6d5ef81b3c3071d0fb", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36411, + "estimated_input_tokens": 9103, + "estimated_eval_tokens": 18462 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a grab bag of unrelated Transformers issues across tokenizers, multimodal processors, generation, loading/offloading, configs, and docs. None of the soft pairs look like the same underlying bug/change, so there is no credible duplicate consolidation here.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out: the items span many distinct subsystems and failure modes, so any representative would be misleading.", + "canonical_pr_reason": null, + "best_issue_reason": "No strong global best issue: even the broadest reports are specific to one model/path and do not represent the cluster as a whole.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "Different bugs: BeitImageProcessorFast label reduction vs MolmoForCausalLM missing all_tied_weights_keys." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Different loading problems: local cache-path collision vs save_pretrained permission propagation." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer failures in different models and code paths (LayoutLMv2 NER padding vs MLuke task AttributeError)." + }, + { + "left": "issue:43425", + "right": "issue:46821", + "accept": false, + "reason": "Torch 2.10 compatibility issue is unrelated to TimesFM window_size AttributeError." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Training early-stopping metric selection vs image processor void-segmentation handling are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Different config/API bugs: Gemma3TextConfig discrepancy vs GPTNeoX rotary_pct reload regression." + }, + { + "left": "issue:45701", + "right": "issue:47145", + "accept": false, + "reason": "Broad tokenization-version behavior vs FP8 compressed-tensors checkpoint loading corruption; not the same defect." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Different models and symptoms: Qwen3Moe FineGrainedFP8Config vs Llava Next segfault." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS CLIP bus error vs generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Tokenizer inconsistency across slow/fast implementations is not the same as T5 add_special_tokens name conflict." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Image segmentation-map processing vs BarkProcessor voice_preset are unrelated processor bugs." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantized offloading behavior vs T5 apex RMSNorm memory leak are different code paths and root causes." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Different multimodal/Qwen bugs: video input ValueError vs weight_only loading error." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Both Qwen-related, but one is sliding_window behavior and the other is from_config text_config unwrapping; different bugs." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategies is not the same as stale docs for removed pipeline tasks." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba is unrelated to BigBirdTokenizer special-token registration." + }, + { + "left": "issue:43452", + "right": "issue:47020", + "accept": false, + "reason": "Different tokenizer/model loading issues: gguf_file handling vs XLMRobertaTokenizer Unigram input type." + }, + { + "left": "issue:44534", + "right": "issue:46665", + "accept": false, + "reason": "Non-persistent buffer junk in v5 vs from_pretrained behavior inside Python classes are different bugs." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3 auto device-mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "CPU offload device mismatch for Bark-small vs BitNet packed-weight unpacking during accelerate loading are different issues." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Local-path trailing slash import warning is unrelated to undefined video-branch variables in multimodal token counting." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token counter not updating vs progressive generation with inputs_embeds/past_key_values are different problems." + }, + { + "left": "issue:43531", + "right": "issue:46821", + "accept": false, + "reason": "Both mention windowing, but Qwen3-MoE sliding_window and TimesFM window_size are different model-specific bugs." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset failure and Qwen3 device-mapping CUDA assert are unrelated." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer in serve is not the same as Qwen3.5 streaming chat/completions failure." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "CPU offload device mismatch and BigBirdTokenizer decode/special-token issues are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "gather_for_metrics label truncation and SAM3 text/bbox odd behavior are different bugs." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug is unrelated to missing Float8 storage support." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "from_config text_config unwrapping and CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding are different 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": 27, + "review_unit_index": 17, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:41720", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43493", + "issue:43526", + "issue:43540", + "issue:43644", + "issue:43653", + "issue:43746", + "issue:43844", + "issue:43873", + "issue:43881", + "issue:43927", + "issue:43994", + "issue:44062", + "issue:44077", + "issue:44368", + "issue:44451", + "issue:44464", + "issue:44514", + "issue:44545", + "issue:44610", + "issue:44625", + "issue:44671", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45357", + "issue:45479", + "issue:45704", + "issue:45759", + "issue:45923", + "issue:46050", + "issue:46402", + "issue:46655", + "issue:46665", + "issue:46693", + "issue:46858", + "issue:47050" + ], + "soft_pairs": [ + "issue:44451|issue:44991", + "issue:43540|issue:46858", + "issue:45357|issue:46402", + "issue:44077|issue:44625", + "issue:43644|issue:46665", + "issue:44987|issue:45042", + "issue:44062|issue:44451", + "issue:44898|issue:45072", + "issue:43381|issue:43844", + "issue:44625|issue:45479", + "issue:39692|issue:44610", + "issue:41720|issue:45704", + "issue:43526|issue:43881", + "issue:44368|issue:45357", + "issue:43381|issue:45704", + "issue:44918|issue:44936", + "issue:43746|issue:46050", + "issue:46655|issue:46693", + "issue:43421|issue:43927", + "issue:45759|issue:45923", + "issue:45704|issue:45923", + "issue:43746|issue:43873", + "issue:42886|issue:44843", + "issue:30064|issue:36010", + "issue:44514|issue:44545", + "issue:39692|issue:43493", + "issue:43421|issue:43653", + "issue:43994|issue:44671", + "issue:44464|issue:47050" + ], + "prepared_review_unit_hash": "abdbc1d57cb7011500813d29b33202fcacf519244a878b0461dcddad19130f38", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36678, + "estimated_input_tokens": 9170, + "estimated_eval_tokens": 18596 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous issue reports, but two soft-duplicate pairs stand out: the Qwen2_5_VLProcessor batched apply_chat_template crash, and the HF_HUB_OFFLINE tokenizer/model_info offline-mode failure. The rest look like related-but-distinct bugs rather than the same underlying change.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 46693 is the strongest single anchor by activity and impact: it is open, has the highest discussion volume, and describes a broad performance regression affecting a major generation path.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46693 is the best overall representative because it is actively discussed, broadly scoped, and most likely to be the cluster\u2019s main user-facing problem rather than a narrow edge case.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/model loading failures, but they concern different models and likely different root causes; too broad to treat as the same bug." + }, + { + "left": "issue:43540", + "right": "issue:46858", + "accept": false, + "reason": "These cover unrelated failures: a Qwen3OmniMoe video-input ValueError versus a static-cache/torch.compile tensor arithmetic bug." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Different model families and different failure modes: incorrect visual encoder keys on save versus missing text encoder weight loading." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "One is about an optional post_init policy in patchtsmixer; the other is config propagation of num_labels in Qwen3.5. Not the same code path." + }, + { + "left": "issue:43644", + "right": "issue:46665", + "accept": false, + "reason": "A non-persistent-buffer initialization bug and an inconsistent from_pretrained-in-Python-classes behavior are too different to merge." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Different underlying issues: a model loading failure for one repo versus PIL image processors wrongly requiring torchvision." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken keyword duplication and a model-loading failure for ScandiBERT are unrelated." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "One is a Perceiver image-size/interpolation issue; the other is a bfloat16 dtype mismatch in inference." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Both mention gradients, but one is about eval-mode checkpointing usage and the other about a ZeRO-3 training regression." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "These are separate classification-config bugs: num_labels propagation versus degenerate zero-loss behavior for num_labels=1." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Both involve vision models, but one is a SigLIP2 doc/example problem and the other is an OmDet-Turbo processor size mismatch." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "Completely different problems: Qwen3 auto device mapping crash versus a T5/apex RMSNorm memory leak." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "A BeitImageProcessorFast label-reduction bug and a glm-4v-9b loading failure are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "A configuration warning about tie_word_embeddings is not the same bug as incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Different layers of the stack: eval-mode checkpointing restriction versus a T5 RMSNorm memory leak." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Both are trainer-related, but one is an embedding unpacking failure with TRL SFT and the other is evaluate-after-train behavior; no shared concrete bug." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Local checkpoint loading with PEFT adapters and quantized save/load breakage are distinct persistence issues." + }, + { + "left": "issue:46655", + "right": "issue:46693", + "accept": false, + "reason": "Both are performance/processor-adjacent, but one is a video padding bug in SmolVLM and the other is a long-context FA2 generation regression." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "A tokenizer post-processor update bug and a DiaConfig custom-token-ID persistence bug are different state-management problems." + }, + { + "left": "issue:45759", + "right": "issue:45923", + "accept": false, + "reason": "One is about unwrapping text_config in composite multimodal configs; the other is a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "A memory leak in apex FusedRMSNorm is unrelated to a missing capability flag on a trust_remote_code model." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "PEFT local checkpoint loading and quantization/offloading behavior are related only at a broad subsystem level, not the same bug." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Both describe the same offline-mode failure path: tokenizer loading unconditionally triggers hub metadata access/model_info(), which breaks HF_HUB_OFFLINE." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image processor void-segmentation-map handling and a GenerationMixin import error are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "These are the same crash report with effectively identical titles and behavior: Qwen2_5_VLProcessor.apply_chat_template fails on batched input when padding=False." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Both mention SigLIP2, but one is a documentation/example issue and the other is a implementation-vs-JAX discrepancy; not the same concrete fix." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "TokenizerBackend post-processor updates and BigBirdTokenizer special-token registration are distinct tokenizer bugs." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and symptoms: SigLIP2 nonsensical outputs versus CamemBERT masked LM prediction regression." + }, + { + "left": "issue:44464", + "right": "issue:47050", + "accept": false, + "reason": "Chunked generation with compiled forward and fp16 ONNX attention dtype mismatch are separate generation/export problems." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: issue 46693 is a reasonable representative given its high discussion activity and broad performance-regression scope, and the two accepted soft pairs are conservative and well-supported by the titles. The remaining verdicts stay appropriately cautious and do not overstate sameness across clearly different bugs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 18, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:30064", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43295", + "issue:43335", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43611", + "issue:43643", + "issue:43653", + "issue:43720", + "issue:43756", + "issue:43782", + "issue:43819", + "issue:43873", + "issue:43927", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44361", + "issue:44488", + "issue:44625", + "issue:44661", + "issue:44857", + "issue:44898", + "issue:44991", + "issue:44995", + "issue:45081", + "issue:45356", + "issue:45381", + "issue:45834", + "issue:45907", + "issue:45987", + "issue:46032", + "issue:46294", + "issue:46463", + "issue:46665", + "issue:46693", + "issue:46829", + "issue:46832", + "issue:46858", + "issue:47020", + "issue:47050" + ], + "soft_pairs": [ + "issue:46032|issue:46463", + "issue:44857|issue:47050", + "issue:43782|issue:45381", + "issue:45081|issue:45356", + "issue:43421|issue:44186", + "issue:45356|issue:45907", + "issue:44995|issue:46858", + "issue:39692|issue:43873", + "issue:39692|issue:47050", + "issue:43720|issue:43819", + "issue:43295|issue:45834", + "issue:43653|issue:43720", + "issue:43550|issue:43927", + "issue:43550|issue:44361", + "issue:43596|issue:44220", + "issue:46829|issue:46832", + "issue:44062|issue:44488", + "issue:44625|issue:47050", + "issue:43531|issue:43643", + "issue:43540|issue:46294", + "issue:45381|issue:46693", + "issue:45987|issue:46829", + "issue:43611|issue:46665", + "issue:30064|issue:36331", + "issue:41720|issue:44991", + "issue:44220|issue:44661", + "issue:43653|issue:43819", + "issue:44186|issue:44898", + "issue:44488|issue:47020", + "issue:43550|issue:43819", + "issue:43335|issue:43756" + ], + "prepared_review_unit_hash": "1e10afac0c2a64547170327c91b01238ad8020a77deeb73514342cd031634f5f", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36768, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a true duplicate set: it spans many unrelated bug reports across tokenizers, model loading, vision/video inputs, cache behavior, quantization, and configs. None of the soft pairs look like the same underlying code-path failure.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Different subsystems and failures: Mamba2 cache correctness vs T5 decoder-input validation error messaging." + }, + { + "left": "issue:44857", + "right": "issue:47050", + "accept": false, + "reason": "Unrelated models and bug types: AMP/CUDA loss crash vs ONNX fp16 attention dtype mismatch." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen vision-related, but they are different models and different failures (weight_only load error vs wrong video position IDs)." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Tokenizer regressions, but not the same bug: one is a backend_tokenizer attribute error, the other is codec handling/warning regression in a specific tokenizer version." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Special-token post-processing behavior vs LayoutLMv2 NER/padding crash are different code paths." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regression vs HF Hub revision forwarding in CI; no shared underlying bug." + }, + { + "left": "issue:44995", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache/indexing, but one is a stale model cache on second forward and the other is a static-cache sequence-length type issue under torch.compile." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "Documentation/example errors are unrelated to quantization offloading behavior." + }, + { + "left": "issue:39692", + "right": "issue:47050", + "accept": false, + "reason": "Docs/example problems vs Canine ONNX export dtype mismatch are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "Different models and failures: accelerate loading of packed weights vs DAC latent/STE mismatch." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Regression in processor/tokenizer access vs long OCR index error in Kosmos2.5; not the same defect." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBird special-token registration bug is unrelated to BitNet packed-weight loading." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile/SDPA crash in Bamba is unrelated to DiaConfig token ID persistence." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Different model families and unrelated failures: compile/SDPA crash vs tokenizer AttributeError." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Deepspeed ZeRO-3/BertModel indexing vs audio feature extraction issue; no common bug." + }, + { + "left": "issue:46829", + "right": "issue:46832", + "accept": false, + "reason": "Both are output-shape/forward issues, but they affect different models and different invariants." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken keyword conflict vs failure to load a specific tokenizer model are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:47050", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is unrelated to Canine fp16 ONNX export dtype mismatch." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to AutoConfig trust_remote_code field loss." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-related Qwen issues, but the failing path differs: video input processing ValueError vs chat-template feature/token count mismatch." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Wrong video position IDs in Qwen2.5-VL is unrelated to a long-context FA2 performance regression." + }, + { + "left": "issue:45987", + "right": "issue:46829", + "accept": false, + "reason": "StaticCache return type bug is unrelated to VideoPrism hidden_states/output packaging." + }, + { + "left": "issue:43611", + "right": "issue:46665", + "accept": false, + "reason": "Both involve from_pretrained/loading, but one is base_model_prefix breakage in v5 and the other is inconsistent behavior inside Python classes." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image segmentation map processing and CustomTrainer compute_loss signature mismatch are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "Qwen3 device-mapping CUDA assert is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio fbank extraction issue is unrelated to add-new-model-like failing inside TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "Tokenizer special-token decode bug is unrelated to DAC.from_latents/forward mismatch." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes and Perceiver non-default resolution failures are separate model-specific bugs." + }, + { + "left": "issue:44488", + "right": "issue:47020", + "accept": false, + "reason": "Tokenizer loading failure for one model is unrelated to Unigram sequence-vs-dict construction in XLMRobertaTokenizer." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "No shared code path: Bamba compile/SDPA crash vs DAC latent conversion bug." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "Both are model-configuration bugs, but they affect different architectures and different parameters." + } + ] + }, + "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": 27, + "review_unit_index": 19, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:34634", + "issue:36010", + "issue:39692", + "issue:43322", + "issue:43421", + "issue:43540", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43618", + "issue:43643", + "issue:43644", + "issue:43697", + "issue:43749", + "issue:43782", + "issue:43818", + "issue:44008", + "issue:44186", + "issue:44442", + "issue:44448", + "issue:44479", + "issue:44561", + "issue:44671", + "issue:44805", + "issue:44811", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45183", + "issue:45479", + "issue:45584", + "issue:45685", + "issue:45834", + "issue:45987", + "issue:46291", + "issue:46512", + "issue:46858", + "issue:46932", + "issue:47020" + ], + "soft_pairs": [ + "issue:43618|issue:43697", + "issue:45987|issue:46932", + "issue:43643|issue:44877", + "issue:44821|issue:45084", + "issue:44008|issue:46858", + "issue:34634|issue:36010", + "issue:44843|issue:45081", + "issue:39692|issue:45834", + "issue:44186|issue:44442", + "issue:43540|issue:44008", + "issue:44448|issue:44671", + "issue:45584|issue:45834", + "issue:44991|issue:47020", + "issue:45183|issue:45479", + "issue:44991|issue:46291", + "issue:43322|issue:45685", + "issue:43644|issue:43749", + "issue:43606|issue:45072", + "issue:44561|issue:45020", + "issue:43421|issue:44811", + "issue:43550|issue:45072", + "issue:43577|issue:43818", + "issue:44479|issue:45834", + "issue:44805|issue:44936", + "issue:43582|issue:46512", + "issue:43782|issue:44977" + ], + "prepared_review_unit_hash": "4dde6af676c3b3634b0d323c1afb2746219805d953ba3b73fad3c9938251bab2", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35191, + "estimated_input_tokens": 8798, + "estimated_eval_tokens": 17852 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All soft pairs look like false positives: they share broad subsystems or regression themes, but not the same concrete bug/change. No clear canonical issue or PR emerges from this set.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Both are Transformers v5 regressions, but they affect different models and different failures (CLIP attentions vs RTDetrV2 outputs)." + }, + { + "left": "issue:45987", + "right": "issue:46932", + "accept": false, + "reason": "StaticCache seq-length semantics and Gemma4 quantized embedder dtype casting are unrelated code paths." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both concern config loading, but one is missing fields with trust_remote_code and the other is strict granite_speech config validation; not the same bug." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor URL loading and template-node compilation errors are unrelated." + }, + { + "left": "issue:44008", + "right": "issue:46858", + "accept": false, + "reason": "Gemma3n variable-name collision in forward() and cache.get_seq_length() returning a device tensor are different bugs." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset handling and the GenerationMixin import error are unrelated." + }, + { + "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 crash; separate failure modes." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "SigLIP2 doc-example/model mismatch and Kosmos2.5 OCR long-input indexing are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crashes and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are different tokenizer issues." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Qwen3OmniMoe video-processing ValueError and Gemma3n audio mask collision are unrelated multimodal bugs." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Both are v5 output regressions, but they concern different models and different behavior changes." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription generation and Kosmos2.5 long OCR input indexing are not the same problem." + }, + { + "left": "issue:44991", + "right": "issue:47020", + "accept": false, + "reason": "Tokenizer loading for EMBEDDIA/est-roberta and XLMRobertaTokenizer Unigram init are different root causes." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Input validation errors and single-label classification zero-loss behavior are separate sequence-classification issues." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading failure and pipelines import failing on kernels v0.15 are unrelated." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next loading segfault and an MPS histogram dtype limitation are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer junk and FSDP CPU RAM efficient loading are different breakages." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch for Bark and bfloat16 dtype mismatches for SwitchTransformers/TimmWrapperModel are different concrete issues." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Both mention remote_code breakage, but 45020 is a broad umbrella and does not identify the same specific code-path as the is_torch_fx_available removal." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Runtime special-token/post-processor updates and Whisper skip_special_tokens decoding are unrelated tokenizer/processor behaviors." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba-9B-v2 and dtype mismatches in other models are different problems." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "BLIP2 dtype propagation and Video-LLaVA missing temporal attention/shared weights are unrelated multimodal issues." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Qwen video-input regression and Kosmos2.5 OCR long-input indexing are different model failures." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask shape mismatch IndexError and trainer.evaluate() failing after train() do not point to the same bug." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon warmup TypeError and stale config.dtype after from_config(dtype=...) are unrelated dtype/runtime issues." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation issues are 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": 27, + "review_unit_index": 20, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36331", + "issue:42175", + "issue:43262", + "issue:43388", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43645", + "issue:43653", + "issue:43716", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43931", + "issue:43994", + "issue:44008", + "issue:44291", + "issue:44589", + "issue:44610", + "issue:44671", + "issue:44857", + "issue:44912", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45381", + "issue:45701", + "issue:46032", + "issue:46175", + "issue:46285", + "issue:46294", + "issue:46463", + "issue:46665", + "issue:46697", + "issue:46728", + "issue:46846", + "issue:46932", + "issue:47217" + ], + "soft_pairs": [ + "issue:43825|issue:43994", + "issue:43388|issue:46032", + "issue:46294|issue:46463", + "issue:43931|issue:46846", + "issue:43716|issue:44008", + "issue:28218|issue:29159", + "issue:43540|issue:45381", + "issue:44610|issue:45072", + "issue:43720|issue:45072", + "issue:43749|issue:44291", + "issue:43645|issue:46665", + "issue:34634|issue:36331", + "issue:43825|issue:43827", + "issue:42175|issue:45081", + "issue:44671|issue:45701", + "issue:43577|issue:43582", + "issue:45003|issue:46285", + "issue:46175|issue:46932", + "issue:43262|issue:47217", + "issue:46697|issue:46728", + "issue:43582|issue:44857", + "issue:43653|issue:44912", + "issue:43262|issue:43819", + "issue:43782|issue:45325", + "issue:44589|issue:45042", + "issue:43742|issue:45020" + ], + "prepared_review_unit_hash": "c176d9dcfecaa6d9f15b4a460a5d7f99e35ff84ac0afb785dfc37e820df22a0b", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35943, + "estimated_input_tokens": 8986, + "estimated_eval_tokens": 18228 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is largely heterogeneous: the soft pairs are mostly superficial similarities across different models, tokenizers, loaders, or processors rather than the same underlying bug/change. I would not merge any of the proposed pairs.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue emerges; the items span many unrelated bugs and only share broad transformer-like themes.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue for deduplication; the cluster does not center on one underlying defect.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Different problems: a pipeline error-message regression vs. a model producing bad outputs. Same broad area, not the same bug." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Unrelated subsystems: metric gathering vs. Mamba2 cache behavior." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Different validation paths and models; one is Qwen video chat templating, the other is T5 decoder-input error messaging." + }, + { + "left": "issue:43931", + "right": "issue:46846", + "accept": false, + "reason": "Different Qwen model-loading/shape issues with no clear shared code path." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Different multimodal components and failure modes: image-preprocessor dtype mismatch vs. Gemma3n forward variable collision." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Only loosely related tokenizer behavior; one is added-token spacing, the other is a broad slow/fast inconsistency report." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Different models and different video-position/processing bugs." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor-size mismatch and dtype mismatch are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "Packed-weight loading bug is unrelated to bfloat16 dtype mismatch." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Both involve loading, but one is FSDP CPU RAM-efficient loading and the other is init_empty_weights argument compatibility." + }, + { + "left": "issue:43645", + "right": "issue:46665", + "accept": false, + "reason": "Different initialization contexts and failure modes; not the same defect." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Completely different areas: BarkProcessor preset handling vs. trainer loss signature changes." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "One is an error message bug, the other is stale documentation referencing pipeline(); related topic, not one fix." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "TensorFlow backend packaging and tokenizer regex patching are unrelated." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "Tokenization-version changes may be adjacent, but these describe different concrete failures and are not clearly the same bug." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Different dtype/load state issue vs. Apple Silicon allocator TypeError." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "sys.modules access safety and CLIP model structure inconsistency are unrelated." + }, + { + "left": "issue:46175", + "right": "issue:46932", + "accept": false, + "reason": "One is a huge one-hot position embedding memory issue; the other is dtype casting for quantized embedders." + }, + { + "left": "issue:43262", + "right": "issue:47217", + "accept": false, + "reason": "Audio chat-template sampling-rate default and VL image-placeholder handling are different bugs." + }, + { + "left": "issue:46697", + "right": "issue:46728", + "accept": false, + "reason": "Both concern image processing, but the underlying logic differs: over-padding vs. patch-count computation." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError is unrelated to AMP/CUDA loss crashes." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration and MXFP4 weight loading are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio template defaults and DAC forward/from_latents mismatch are unrelated." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen-VL bugs: weight_only loading error vs. incorrect temporal position scaling." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Storage-object lookup failure and torchvision dependency regression are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "KeyError loading one model is not the same as remote_code compatibility breaking across versions." + } + ] + }, + "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": 27, + "review_unit_index": 21, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43638", + "issue:43643", + "issue:43645", + "issue:43696", + "issue:43697", + "issue:43742", + "issue:43746", + "issue:43819", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:44060", + "issue:44368", + "issue:44479", + "issue:44554", + "issue:44610", + "issue:44811", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45081", + "issue:45084", + "issue:45372", + "issue:45406", + "issue:45538", + "issue:45600", + "issue:45759", + "issue:45859", + "issue:46209", + "issue:46291", + "issue:46326", + "issue:47050" + ], + "soft_pairs": [ + "issue:45372|issue:46326", + "issue:43550|issue:44912", + "issue:36010|issue:43645", + "issue:36010|issue:36331", + "issue:43257|issue:43329", + "issue:43531|issue:45042", + "issue:43526|issue:43697", + "issue:43638|issue:44936", + "issue:42175|issue:46291", + "issue:43452|issue:44991", + "issue:43906|issue:46209", + "issue:43819|issue:43927", + "issue:45081|issue:45406", + "issue:44060|issue:44368", + "issue:45084|issue:45538", + "issue:43329|issue:45600", + "issue:43526|issue:44554", + "issue:42490|issue:43257", + "issue:43746|issue:44811", + "issue:43854|issue:44977", + "issue:43643|issue:45759", + "issue:45859|issue:47050", + "issue:38175|issue:44610", + "issue:39692|issue:44912", + "issue:43262|issue:44811", + "issue:43257|issue:43696", + "issue:43329|issue:44479", + "issue:42915|issue:43742" + ], + "prepared_review_unit_hash": "498c64da19d01704364425e7e94036ba32d5419ebd0e1c318b55daec7a9bfcc2", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36549, + "estimated_input_tokens": 9138, + "estimated_eval_tokens": 18532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the soft-similarity pairs are mostly superficial (shared buzzwords like loading, tokenizer, MoE, processor, or model naming) but do not describe the same underlying bug/change. No soft edge should be merged.", + "confidence": 0.95, + "canonical_issue_reason": "No true duplicate center emerges from this set; if one issue must anchor the cluster, issue 43257 is the most substantive and well-scoped among the Qwen/MoE-loading reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43257 is the clearest, most concrete bug report in the set, but it is only a local anchor, not a universal representative of the whole heterogeneous cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "Both involve mistral_common-related tokenizers, but one is an import error in Gemma 4 processor loading and the other is a continue_final_message behavior bug; different failure modes and code paths." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "Both mention Qwen/OSS models and quantization/attention, but one is a torch.compile + SDPA failure and the other is an MXFP4 loading fallback; not the same bug." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import error vs Jupyter custom-model initialization breakage in Transformers 5.0; unrelated symptoms and fixes." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "Importing GenerationMixin and CustomTrainer.compute_loss signature changes are unrelated API breakages." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Both are Qwen3 MoE-related, but one is about accelerate+deepspeed weight conversion and the other is a multimodal token-counting video-branch NameError/undefined variable bug." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Both concern processor/input handling, but one is a Qwen3-MoE sliding_window issue and the other is a PIL backend dependency regression on torchvision." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output drift are different model-specific bugs with different code paths." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Deepspeed ZeRO-3 index error during training and trainer.evaluate() failing after train() are distinct trainer/runtime issues." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Torch backend installation vs kernels v0.15 import/version constraint is a dependency-management mismatch, not the same issue." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/model loading problems, but gguf_file support failure and a specific tokenizer load regression for EMBEDDIA/est-roberta are not the same bug." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "A reproduction of an earlier issue and an FP8 inference-on-B200 bug are unrelated." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC latent mismatch and DiaConfig custom-token-ID loss are different model/config correctness issues." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Both hit Gemma/Mistral tokenizer-related loading paths, but one is a Mistral regex patch crash and the other is a Gemma4Processor missing _tokenizer in transformers serve." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both involve Qwen3.x MoE tied-weight warnings, but one is a specific incorrect weight-tying warning and the other is a LoRA tuning warning about tie_word_embeddings; not the same defect." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template compilation error and CLIPTokenizer max_length constant are unrelated tokenizer/template issues." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Video token counting bug and auto_mappings.py CI break from removed configs are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Beit label reduction bug and an MPS attention correctness issue have no shared code path." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Loading a local model path with a trailing slash and Qwen3 MoE weight conversion under accelerate+deepspeed are unrelated loader behaviors." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT adapter local checkpoint loading and Whisper processor.batch_decode skipping special tokens are different processor/model bugs." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Unit-test model loading for zai-org/GLM-4.7-Flash and Qwen3.5 flash-attention generation failure are separate model-specific issues." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "trust_remote_code AutoConfig field loss and AutoModelForCausalLM.from_config not unwrapping text_config are distinct config-loading bugs." + }, + { + "left": "issue:45859", + "right": "issue:47050", + "accept": false, + "reason": "CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding and Canine fp16 ONNX export dtype mismatch are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "SigLIP2 zero probabilities and OmDet-Turbo processor/model input-size mismatch are different vision-model issues." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "SigLIP2 documentation example errors and git-oss-20b MXFP4 loading fallback are unrelated; one is docs/example correctness, the other quantization behavior." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio apply_chat_template default sample-rate bug and Whisper batch_decode skip_special_tokens bug affect different processor methods." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and GPT-oss-20b CUDA OOM are not the same failure." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Undefined multimodal token helper variables in the video branch and a broader v5.3.0 video-input regression across multiple VLMs are related only by domain, not by concrete bug." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure for Qwen3Moe and KeyError loading facebook/MobileLLM-125M are separate loader/config 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": 27, + "review_unit_index": 22, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28282", + "issue:29159", + "issue:29870", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:39379", + "issue:40990", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43329", + "issue:43425", + "issue:43450", + "issue:43454", + "issue:43525", + "issue:43540", + "issue:43577", + "issue:43650", + "issue:43697", + "issue:43749", + "issue:43824", + "issue:43906", + "issue:43931", + "issue:44291", + "issue:44387", + "issue:44554", + "issue:44568", + "issue:44617", + "issue:44756", + "issue:44821", + "issue:44857", + "issue:44912", + "issue:44987", + "issue:45042", + "issue:45137", + "issue:45356", + "issue:45357", + "issue:45405", + "issue:45459", + "issue:45538", + "issue:45803", + "issue:46291", + "issue:46421", + "issue:46424", + "issue:46701", + "issue:46832", + "issue:47145" + ], + "soft_pairs": [ + "issue:43257|issue:43931", + "issue:43329|issue:43450", + "issue:44291|issue:45042", + "issue:34567|issue:40990", + "issue:44912|issue:47145", + "issue:42915|issue:44821", + "issue:44387|issue:44617", + "issue:29159|issue:34567", + "issue:45459|issue:45803", + "issue:46421|issue:46424", + "issue:43525|issue:46291", + "issue:43650|issue:43906", + "issue:28282|issue:29870", + "issue:44987|issue:45538", + "issue:43425|issue:46291", + "issue:43749|issue:44756", + "issue:28282|issue:34689", + "issue:43540|issue:46701", + "issue:44568|issue:45356", + "issue:43454|issue:43577", + "issue:33357|issue:34567", + "issue:42175|issue:43824", + "issue:39379|issue:44554", + "issue:43697|issue:45357", + "issue:45137|issue:45405", + "issue:44857|issue:46832" + ], + "prepared_review_unit_hash": "a33381f07fee0f46fc75d93cf74aa210c67e46d28c6377c1053a3efce52b3ab7", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35518, + "estimated_input_tokens": 8880, + "estimated_eval_tokens": 18016 + }, + "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 broad keywords (tokenizer, video, quantization, cache, loading) but describe different failures and code paths, so none are safe duplicate merges.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Both mention Qwen3 model loading, but one is about MoE weight conversion with accelerate/deepspeed and the other is a Qwen3-VL shape-mismatch load error." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Both are video-related, but one is an undefined-variable bug in multimodal token counting and the other is a batched video-processor shape bug." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Different code paths: init_empty_weights argument handling versus PIL backend processors wrongly requiring torchvision." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token counting and high perplexity on a model are unrelated symptoms and subsystems." + }, + { + "left": "issue:44912", + "right": "issue:47145", + "accept": false, + "reason": "Both involve quantized checkpoints, but one is MXFP4 falling back to bf16 and the other is FP8 weight_scale not being applied; different formats and failures." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to loading AutoImageProcessor from a URL." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both are OOM reports, but they come from different models and causes: int4 reserved-memory growth versus Sam3Video memory use." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency is a different bug from a TrainerState counter not updating." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Both involve exception handling, but they are different helpers and different swallowed-error behavior." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both touch cache logic, but one is lazy-init recompilation with chunked prefill and the other is max_cache_length being ignored outside quantized cache." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Llama4Config missing pad_token_id is unrelated to kernels import/version requirements in pipelines." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "No substantive duplicate signal: one is a vague placeholder issue and the other is a reproduction of a different upstream bug." + }, + { + "left": "issue:28282", + "right": "issue:29870", + "accept": false, + "reason": "PyTorch missing ImportError and Bart generate decoder_attention_mask UnboundLocalError are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Different issues: a load regression for physical-intelligence/fast versus CLIPTokenizer's model_max_length constant." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch 2.10 compatibility and kernels import/version enforcement are different dependency problems." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and disabling mmap on Strix Halo are not the same bug or code path." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Missing PyTorch import and Llama 3.2 90B Vision model-loading regression are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46701", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and DiffusionGemma pipeline failure are different model/pipeline bugs." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Different tokenizer regressions: mdeberta add_special_tokens behavior versus Kimi-K2.5 codec handling/warning issues." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "AyaVision weight tying causing bad generations is unrelated to BLIP2 dtype remaining float32." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS bus error in CLIP loading and a TrainerState token counter bug do not overlap." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Missing tensorflow backend in a torch install is unrelated to importing a missing Qwen2_5_VL class." + }, + { + "left": "issue:39379", + "right": "issue:44554", + "accept": false, + "reason": "Both are correctness bugs, but one is an mllama mask logic issue and the other is an MPS attention backend issue with different head dimensions." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "RT-DETR output changes and Qwen3.5 save_pretrained visual-key corruption are separate regressions." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque failure and a PEFT version bump are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:46832", + "accept": false, + "reason": "Different RT-DETR/LwDetr bugs with different failure modes: AMP crash versus feature-level mismatch TypeError." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 23, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:24643", + "issue:27957", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:34567", + "issue:36331", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43404", + "issue:43452", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44568", + "issue:44589", + "issue:44792", + "issue:44821", + "issue:44849", + "issue:44898", + "issue:44936", + "issue:44964", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45081", + "issue:45127", + "issue:45381", + "issue:45488", + "issue:45538", + "issue:45701", + "issue:45907", + "issue:46512", + "issue:46693", + "issue:46821", + "issue:46869", + "issue:47020" + ], + "soft_pairs": [ + "issue:43404|issue:45127", + "issue:45081|issue:47020", + "issue:28282|issue:29127", + "issue:44936|issue:46821", + "issue:43742|issue:44291", + "issue:42915|issue:45538", + "issue:44479|issue:46693", + "issue:45538|issue:46869", + "issue:44062|issue:44991", + "issue:24643|issue:29159", + "issue:27957|issue:28282", + "issue:43452|issue:43582", + "issue:42915|issue:43749", + "issue:43577|issue:46512", + "issue:44849|issue:45701", + "issue:41720|issue:43540", + "issue:42175|issue:44488", + "issue:43540|issue:44560", + "issue:44291|issue:44821", + "issue:44987|issue:45907", + "issue:43540|issue:44479", + "issue:34567|issue:36331", + "issue:44479|issue:45381", + "issue:44792|issue:44964", + "issue:38175|issue:44898", + "issue:44062|issue:45081", + "issue:44568|issue:45488", + "issue:45042|issue:45081", + "issue:43257|issue:43756", + "issue:43742|issue:44589", + "issue:43582|issue:46821" + ], + "prepared_review_unit_hash": "062a482a16f8e9db64c225ad8389612fb342cb2c92dd13bb953e8d117937c344", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37011, + "estimated_input_tokens": 9253, + "estimated_eval_tokens": 18762 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: tokenizer regressions, model-loading failures, training bugs, and multimodal/video issues. Most soft pairs are not true duplicates. The only near-duplicate-looking subgroup is the Qwen video-input regression pair.", + "confidence": 0.56, + "canonical_issue_reason": "issue:44479 is the clearest umbrella regression report among the items, covering multiple Qwen video-input failures in v5.3.0 and plausibly subsuming the more specific Qwen2.5-VL symptom in 45381.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45488 is the strongest overall issue anchor for the cluster: it is detailed, actively discussed, has inbound references, and describes a broad tokenizer regression with real downstream impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both involve tied/loaded weights, but they are different bugs: Mistral3 lm_head tying vs LoRA merge with extended vocabulary." + }, + { + "left": "issue:45081", + "right": "issue:47020", + "accept": false, + "reason": "Different tokenizer code paths and failures: Mistral regex patch crash vs XLMRoberta Unigram init type error." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Unrelated problems: missing PyTorch import vs a LayoutLMv3 error-message clarity issue." + }, + { + "left": "issue:44936", + "right": "issue:46821", + "accept": false, + "reason": "Trainer lifecycle bug vs TimesFM window_size AttributeError; no shared concrete code path." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Both are loading-time errors, but the causes differ: MobileLLM key error vs init_empty_weights unexpected argument." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "FineGrainedFP8/Qwen3Moe failure is unrelated to CLIPTokenizer model_max_length behavior." + }, + { + "left": "issue:44479", + "right": "issue:46693", + "accept": false, + "reason": "Video-input regression vs FA2 long-context generation performance regression; different bug class." + }, + { + "left": "issue:45538", + "right": "issue:46869", + "accept": false, + "reason": "Tokenizer max-length constant issue vs multi-GPU YOLO device-map failure." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer failures: AddedToken keyword collision vs loading EMBEDDIA/est-roberta." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed training weight-shape runtime error is unrelated to slow/fast tokenizer inconsistency." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention 2 on XLMRoberta is unrelated to AutoModel missing-PyTorch import handling." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "GGUF loading breakage and AppleSilicon caching allocator TypeError are different subsystems." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "Qwen3Moe FP8 config failure does not match the FSDP CPU RAM efficient loading regression." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Both concern dtype propagation, but one is Blip2 loading state and the other is from_config stale config metadata; not the same concrete bug." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-states bug is unrelated to a broad tokenization-change report." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Different Qwen issues: auto device mapping cuda assert vs video-input processing failure." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Packaging/backend dependency issue is unrelated to a model-loading failure for cjvt/sleng-bert." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-related, but they affect different models and the evidence does not show the same root bug." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Unexpected init_empty_weights argument and loading AutoImageProcessor from URL are distinct failures." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Model loading failure for a repo is unrelated to list_repo_files not forwarding revision." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Same general video area, but the model family and failure mode are different enough that the shared root bug is not established." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token counter bug is unrelated to CustomTrainer.compute_loss signature changes." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Both describe the same Qwen video-input regression in the 5.3.0 timeframe; 45381 looks like a specific symptom of the broader 44479 bug." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Janus image-generation test failure and Phi-4 multimodal loading failure are different issues." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 is not the same as Perceiver image-classification resolution handling." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "AddedToken special-argument collision and Mistral regex patch crash are separate tokenizer bugs." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but one is add_special_tokens BOS/EOS behavior and the other is LlamaTokenizer pre-tokenizer override; different code paths." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "PIL backend requiring torchvision is unrelated to the Mistral tokenizer regex crash." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MOE fine-grained FP8 loading and Smollm3 RoPE-layer dropping are unrelated model-specific bugs." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "MobileLLM key error and missing Float8 storage type are different loading/runtime failures." + }, + { + "left": "issue:43582", + "right": "issue:46821", + "accept": false, + "reason": "AppleSilicon allocator warmup TypeError is unrelated to the TimesFM window_size AttributeError." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The only accepted pair, 44479/45381, is plausibly the same Qwen video-input regression family, while the rest are reasonably rejected as distinct bugs. The chosen anchors are also supported by the activity/inbound-reference signals 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": 27, + "review_unit_index": 24, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29870", + "issue:32090", + "issue:41762", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43577", + "issue:43697", + "issue:43716", + "issue:43720", + "issue:43756", + "issue:43906", + "issue:44488", + "issue:44568", + "issue:44625", + "issue:44737", + "issue:44821", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44936", + "issue:45072", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45405", + "issue:45479", + "issue:45588", + "issue:45600", + "issue:45663", + "issue:45672", + "issue:45674", + "issue:45823", + "issue:45920", + "issue:46121", + "issue:46399", + "issue:46463", + "issue:46468", + "issue:46491", + "issue:46693", + "issue:46701", + "issue:46829", + "issue:46846", + "issue:46899" + ], + "soft_pairs": [ + "issue:43531|issue:44821", + "issue:43756|issue:46399", + "issue:43756|issue:45325", + "issue:45405|issue:45600", + "issue:44737|issue:46463", + "issue:43716|issue:46701", + "issue:44568|issue:46491", + "issue:45823|issue:46463", + "issue:24643|issue:29870", + "issue:43577|issue:45072", + "issue:43906|issue:46693", + "issue:45672|issue:45674", + "issue:44625|issue:44877", + "issue:43452|issue:44488", + "issue:43540|issue:46693", + "issue:28282|issue:32090", + "issue:43906|issue:46468", + "issue:43697|issue:46829", + "issue:44913|issue:46121", + "issue:45663|issue:46846", + "issue:44737|issue:45588", + "issue:43526|issue:45479", + "issue:46463|issue:46899", + "issue:43697|issue:45216", + "issue:41762|issue:44936", + "issue:45357|issue:45920", + "issue:43720|issue:44898" + ], + "prepared_review_unit_hash": "236fc104afbc0e918412f6cd5798e401f1e22aec75e9d1c973017af6aa527658", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35864, + "estimated_input_tokens": 8966, + "estimated_eval_tokens": 18188 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These soft matches are mostly keyword-level overlaps (dtype, load, config, generation) across unrelated models and code paths. None of the candidate pairs look like the same underlying bug/change, so there is no clear canonical issue in this cluster.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue: the items are heterogeneous and the soft-similarity pairs do not converge on one shared defect.", + "canonical_pr_reason": null, + "best_issue_reason": "No strong global best issue: none of these issues is representative enough to serve as a reliable cluster anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Different features and failures: Qwen3-MoE sliding-window behavior vs loading AutoImageProcessor from a URL." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "Both mention model internals, but the bugs are unrelated: Smollm3 RoPE-layer dropping vs Qwen3.5 kernel registration failure." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different models and different bugs: RoPE-layer selection in Smollm3 vs Qwen2.5-VL temporal position_id scaling." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "A dependency-version bump is unrelated to removing stale auto-mapping entries; no shared code-path bug." + }, + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "XLNet CPU tensor creation bug vs T5 decoder-input validation error message; unrelated subsystems." + }, + { + "left": "issue:43716", + "right": "issue:46701", + "accept": false, + "reason": "Different multimodal pipelines and failure modes: dtype mismatch in Mistral-3 image preprocessing vs broken DiffusionGemma image-text-to-text generation." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Tokenizer behavior differs: one is missing BOS/EOS insertion, the other is a leading-space round-trip regression in CodeLlama." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 device-placement mismatch during forward is unrelated to T5 decoder-input validation." + }, + { + "left": "issue:24643", + "right": "issue:29870", + "accept": false, + "reason": "DeepSpeed training weight-shape error and Bart generate() unbound local variable are different bugs in different execution paths." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both are dtype-related, but they affect different models and causes: Blip2 loading dtypes vs SwitchTransformers/TimmWrapper bfloat16 inference mismatches." + }, + { + "left": "issue:43906", + "right": "issue:46693", + "accept": false, + "reason": "An isolated reproduction of an older issue is not the same as a FA2 long-context performance regression." + }, + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Both involve quantization, but one is an unsafe dtype cast in Gemma4 audio/vision embedders and the other is a BitsAndBytes skip-module exclusion-list bug." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and strict granite_speech config loading are different configuration bugs." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading failures, but the affected objects and root causes differ: gguf_file handling vs cjvt/sleng-bert model loading." + }, + { + "left": "issue:43540", + "right": "issue:46693", + "accept": false, + "reason": "Video input processing in Qwen3OmniMoe is unrelated to FA2 long-context generation performance regression." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "PyTorch import failure and _gpu_broadcast_one NoneType error are unrelated environment/runtime failures." + }, + { + "left": "issue:43906", + "right": "issue:46468", + "accept": false, + "reason": "The Nemotron example breakage is not the same as the isolated reproduction of issue 38071." + }, + { + "left": "issue:43697", + "right": "issue:46829", + "accept": false, + "reason": "RTDetrV2 output differences and VideoPrism hidden_states misassignment are unrelated model-specific bugs." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Both involve rope-related config handling, but one is rotary_pct not persisting on reload and the other is a TypeError from JSON-loaded ignore_keys_at_rope_validation." + }, + { + "left": "issue:45663", + "right": "issue:46846", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states rebuilding and Qwen3.5 tensor-parallel plan omissions are different distributed-training bugs." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU placement bug and flash_attention s_aux=None crash are unrelated code paths." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and degenerate zero-loss sequence classification are different issues in different components." + }, + { + "left": "issue:46463", + "right": "issue:46899", + "accept": false, + "reason": "T5 decoder-input validation messaging is unrelated to Gemma4 4-bit audio/vision dtype casting." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "Different regressions: RTDetrV2 output drift vs Qwen3.5 incorrect save_pretrained checkpoint contents." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading index error is unrelated to trainer.evaluate() failing after train()." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Qwen3.5 visual-encoder checkpoint saving and tokenizer token-ID regressions are different bugs." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific 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": 27, + "review_unit_index": 25, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:29159", + "issue:33357", + "issue:38175", + "issue:43295", + "issue:43452", + "issue:43525", + "issue:43531", + "issue:43643", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43761", + "issue:43819", + "issue:43927", + "issue:43994", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44466", + "issue:44493", + "issue:44568", + "issue:44743", + "issue:44811", + "issue:44843", + "issue:44857", + "issue:44898", + "issue:45005", + "issue:45200", + "issue:45356", + "issue:45372", + "issue:45375", + "issue:45542", + "issue:45701", + "issue:45704", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46402", + "issue:46463", + "issue:46701", + "issue:46829", + "issue:46832", + "issue:46932", + "issue:47000" + ], + "soft_pairs": [ + "issue:44466|issue:45005", + "issue:45200|issue:46463", + "issue:43643|issue:45375", + "issue:44493|issue:46050", + "issue:44361|issue:44442", + "issue:43688|issue:43749", + "issue:43653|issue:44361", + "issue:38175|issue:43295", + "issue:43452|issue:44843", + "issue:43531|issue:43742", + "issue:43819|issue:44811", + "issue:43994|issue:44898", + "issue:43761|issue:46829", + "issue:44857|issue:47000", + "issue:29159|issue:33357", + "issue:45372|issue:46170", + "issue:43525|issue:46701", + "issue:45542|issue:46178", + "issue:46832|issue:46932", + "issue:43688|issue:45704", + "issue:43295|issue:45356", + "issue:43688|issue:45701", + "issue:43653|issue:44568", + "issue:43756|issue:46402", + "issue:43646|issue:43749", + "issue:44464|issue:44743", + "issue:43653|issue:43927" + ], + "prepared_review_unit_hash": "30c1f592a48e4e10447b4411f643577d1ac5a887a27c06f5e8081b95c8d05f44", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35483, + "estimated_input_tokens": 8871, + "estimated_eval_tokens": 17998 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous, and none of the soft pairs look like the same underlying bug with enough confidence to merge. I would keep the issues separate; there is also no PR content here to evaluate.", + "confidence": 0.84, + "canonical_issue_reason": "No single canonical issue cleanly represents the cluster: most items are unrelated regressions or model-specific bugs. If an anchor is needed, 44466 is the closest because it is the most detailed tied-weights report.", + "canonical_pr_reason": null, + "best_issue_reason": "44466 is the strongest standalone report in the set, but it only covers a tied-weight serialization regression and does not subsume the rest of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights in v5, but one is a device-dependent serialization issue and the other is a broader translation-model complaint; not clearly the same bug." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Different failures: Gemma 4 mm_token_type_ids behavior vs T5 decoder-input validation messaging." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Both involve config fields being missing, but the causes differ: trust_remote_code AutoConfig behavior vs @strict dropping a specific Qwen field." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Different symptoms and paths: unexpected position-id keys vs quantized model save/load breakage." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both are tokenizer load errors, but they affect different tokenizers and likely different code paths." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization in MoE models is unrelated to an FSDP CPU RAM-efficient loading regression." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "BigBird special-token registration is unrelated to MLukeTokenizer task AttributeErrors." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "A zero-probability SigLIP model output is not the same bug as a processor/tokenizer regression in custom model code." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both are loading-related, but gguf_file breakage and an unconditional model_info() call in offline mode are distinct issues." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Different model families and failures: Qwen3-MoE sliding_window handling vs a key error loading MobileLLM-125M." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latent reconstruction mismatch and Whisper batch_decode skip_special_tokens handling are unrelated code paths." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Both are vision-model regressions, but one is nonsensical SigLIP outputs and the other is Perceiver resolution/interpolation failure." + }, + { + "left": "issue:43761", + "right": "issue:46829", + "accept": false, + "reason": "Both mention hidden_states, but the models and the failure modes differ: None returned vs wrong tensor placed in the output." + }, + { + "left": "issue:44857", + "right": "issue:47000", + "accept": false, + "reason": "Different losses and different crashes: float16 AMP failure in LwDetrImageLoss vs infeasible cost matrix in RTDetrLoss." + }, + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer inconsistency is not the same as a MacOS bus error in CLIP model loading." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Separate import/dependency problems: Gemma 4 processor loading vs Windows ROCm torch.distributed dependency issues." + }, + { + "left": "issue:43525", + "right": "issue:46701", + "accept": false, + "reason": "Llama4Config missing pad_token_id and DiffusionGemma pipeline failure are unrelated." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both are backend/import issues, but they involve different backends and different failure mechanisms." + }, + { + "left": "issue:46832", + "right": "issue:46932", + "accept": false, + "reason": "RTDetrModel feature-level mismatch and Gemma4 quantized embedder dtype casting are unrelated bugs." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Aux-loss normalization in MoE models is unrelated to a T5/apex memory leak." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor/tokenizer image-handling regression and Kimi-K2.5 codec/mistral-regex regression are different issues." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Aux-loss normalization has nothing to do with a report that transformers version changes tokenization." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "BigBird mask-token registration and mdeberta-v3 BOS/EOS insertion are separate tokenizer-special-token bugs." + }, + { + "left": "issue:43756", + "right": "issue:46402", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and Sam3Model text-encoder weight loading are unrelated model-specific regressions." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model initialization breaks under v5, while the other report is about FSDP CPU RAM-efficient loading." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and Qwen3_5MoE recurrent-state resetting are different code paths." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token registration and DiaConfig custom token ID persistence are unrelated tokenizer/config persistence 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": 27, + "review_unit_index": 26, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:32090", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42898", + "issue:43278", + "issue:43452", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43611", + "issue:43643", + "issue:43688", + "issue:43720", + "issue:43761", + "issue:43782", + "issue:43927", + "issue:44361", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44484", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44625", + "issue:44704", + "issue:44734", + "issue:44821", + "issue:44843", + "issue:44964", + "issue:44977", + "issue:44991", + "issue:45405", + "issue:45440", + "issue:45491", + "issue:46002", + "issue:46050", + "issue:46491", + "issue:46612", + "issue:46693", + "issue:46821", + "issue:46846", + "issue:47013" + ], + "soft_pairs": [ + "issue:44821|issue:46821", + "issue:42898|issue:46491", + "issue:44843|issue:44991", + "issue:44479|issue:44625", + "issue:43452|issue:43577", + "issue:33666|issue:36010", + "issue:43761|issue:47013", + "issue:32090|issue:33666", + "issue:44466|issue:46050", + "issue:44484|issue:44704", + "issue:43688|issue:45440", + "issue:41720|issue:44464", + "issue:44977|issue:46693", + "issue:44464|issue:44977", + "issue:44734|issue:46612", + "issue:43504|issue:43550", + "issue:33666|issue:36331", + "issue:43278|issue:43688", + "issue:43504|issue:43720", + "issue:43782|issue:46693", + "issue:43720|issue:43927", + "issue:43577|issue:44554", + "issue:43606|issue:43927", + "issue:43611|issue:46846", + "issue:43525|issue:43572", + "issue:43504|issue:44610", + "issue:43720|issue:44361", + "issue:45491|issue:46002", + "issue:43643|issue:44561", + "issue:44964|issue:45405" + ], + "prepared_review_unit_hash": "16fda82bb35510e7217b43b4dd84c6936326b0c17afb3d71f0e4cc622953242d", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36847, + "estimated_input_tokens": 9212, + "estimated_eval_tokens": 18680 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: it mixes unrelated tokenizer, config, loading, training, generation, and performance regressions across different models and code paths. None of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue stands out: the items span unrelated model-specific regressions and API breakages rather than one shared defect.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue: there isn\u2019t a representative issue that subsumes the others, and the apparent similarity pairs still break on different code paths and causes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44821", + "right": "issue:46821", + "accept": false, + "reason": "Different failures: URL loading for `AutoImageProcessor` vs a TimesFM `window_size` attribute error." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "Both tokenizer-related, but one is a v5 `clean_up_tokenization_spaces` behavior change and the other is a CodeLlama round-trip spacing regression." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "One is an offline-mode break from `_patch_mistral_regex` calling `model_info()`, the other is a tokenizer load failure for a different model." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Qwen video-input regression vs missing `num_labels` propagation; different models and different bugs." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "`gguf_file` loading breakage vs BLIP2 dtype drift during load; unrelated code paths." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU training issue vs `GenerationMixin` import error; not the same defect." + }, + { + "left": "issue:43761", + "right": "issue:47013", + "accept": false, + "reason": "CLIP hidden-states regression vs xLSTM train-mode tuple unpacking crash; different models and mechanics." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Both involve training, but one is `_gpu_broadcast_one` on `NoneType` and the other is a Qwen2-VL multi-GPU training report." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Different serialization bugs: tied-weight serialization varies by device vs quantized `sinq` save/load breakage." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "`max_shard_size` default question vs `AutoProcessor.from_pretrained` not forwarding kwargs to `cached_file`." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both MoE-related, but one is auxiliary-loss normalization and the other is native vs remote DeepSeekV3 implementation divergence." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "CUDA assert with auto device mapping vs chunked-generation inconsistency under compiled forward; different symptoms and paths." + }, + { + "left": "issue:44977", + "right": "issue:46693", + "accept": false, + "reason": "Qwen3.5 flash-attention correctness issue vs FA2 long-context performance regression; not the same bug." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Compiled-forward chunked generation bug vs flash-attention generation failure; different triggers and code paths." + }, + { + "left": "issue:44734", + "right": "issue:46612", + "accept": false, + "reason": "KV-cache continuation indexing bug in serve vs beam-search cache reorder being skipped for several recurrent models." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Legacy-field model loading failure vs Bamba SDPA/torch.compile failure; unrelated model behaviors." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Both touch training, but one is multi-GPU setup and the other is a `compute_loss()` signature incompatibility." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Embedding dtype drift in eval vs MoE auxiliary-loss normalization; no shared underlying bug." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Beit legacy-field loading issue vs BitNet packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43782", + "right": "issue:46693", + "accept": false, + "reason": "Qwen3VL `weight_only=True` load error vs FA2 performance regression; different failure classes." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "BitNet accelerate-loading bug vs DiaConfig token-ID loss after save/load; unrelated." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "BLIP2 dtype propagation vs MPS attention correctness when value/query head dims differ." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU-offload device mismatch vs config token-ID loss and generation `IndexError`; different code paths." + }, + { + "left": "issue:43611", + "right": "issue:46846", + "accept": false, + "reason": "Base-model-prefix loading regression vs Qwen3.5 TP-plan missing `linear_attn` entries." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Similar theme, but different model configs and different missing fields (`pad_token_id` vs `pad_token_idx`)." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Legacy-field loading failure vs OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Packed-weight loading issue vs MLukeTokenizer task-time AttributeError." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Gemma3 mixed-length GPU NaNs vs classification-head NaNs after checkout/init; different conditions and subsystems." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "`AutoConfig` missing fields under `trust_remote_code` vs removal of `is_torch_fx_available` breaking remote-code models." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Phi-4 multimodal loading failure vs an unreleased PEFT version pin; 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": 27, + "review_unit_index": 27, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28218", + "issue:33357", + "issue:33666", + "issue:39692", + "issue:42796", + "issue:43278", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43572", + "issue:43575", + "issue:43646", + "issue:43701", + "issue:43824", + "issue:43825", + "issue:43844", + "issue:43927", + "issue:43994", + "issue:44291", + "issue:44361", + "issue:44451", + "issue:44466", + "issue:44488", + "issue:44514", + "issue:44554", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44792", + "issue:44843", + "issue:44849", + "issue:44977", + "issue:45290", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45584", + "issue:45812", + "issue:45823", + "issue:46050", + "issue:46192", + "issue:46209", + "issue:46365", + "issue:46489", + "issue:46710", + "issue:46821", + "issue:46832" + ], + "soft_pairs": [ + "issue:44625|issue:45584", + "issue:44514|issue:45290", + "issue:43493|issue:43825", + "issue:44977|issue:46209", + "issue:28218|issue:33357", + "issue:44554|issue:46365", + "issue:43452|issue:44451", + "issue:45823|issue:46050", + "issue:43994|issue:46192", + "issue:42796|issue:44977", + "issue:39692|issue:43295", + "issue:43493|issue:43844", + "issue:43278|issue:43844", + "issue:43278|issue:43701", + "issue:43646|issue:44466", + "issue:46209|issue:46710", + "issue:44568|issue:46489", + "issue:43927|issue:44361", + "issue:33357|issue:33666", + "issue:44849|issue:46209", + "issue:44792|issue:45405", + "issue:43575|issue:45405", + "issue:44568|issue:45812", + "issue:43572|issue:44291", + "issue:44488|issue:44843", + "issue:43824|issue:44589", + "issue:44589|issue:45520", + "issue:46821|issue:46832", + "issue:44849|issue:45357" + ], + "prepared_review_unit_hash": "698d73496e2cdb96d7c67edd37b9dee402dfb49a7c66b67225999c62bb3dcd53", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35882, + "estimated_input_tokens": 8971, + "estimated_eval_tokens": 18198 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are unrelated model, tokenizer, config, or infrastructure bugs. None of the proposed soft pairs look like the same underlying issue, so all soft edges are rejected. If one issue must represent the cluster, the broad Granite tokenizer regression is the strongest candidate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45812 is the strongest representative: it describes a broad, concrete tokenizer regression with clear user impact and inbound references, whereas the other issues are mostly narrow and unrelated.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the best issue candidate because it is a high-signal, broadly scoped tokenizer regression with a concrete failure mode and visible downstream impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Different bugs: Qwen3.5 num_labels propagation vs Whisper empty transcription after token alignment." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Same API area, but different crashes: batched apply_chat_template with padding=False vs assistant tool-call messages with no content." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "Both mention model/pipeline behavior, but one is a SigLIP2 implementation discrepancy and the other is a pipeline translation error message issue." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug vs FP8 inference failure on B200; different code paths and symptoms." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer spacing issue and a MacOS CLIP bus error are unrelated failures." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "MPS attention correctness regression vs Trainer import failing because of datasets; unrelated." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "Both are loading problems, but gguf_file handling and loading a specific model are distinct root causes." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Gemma4 device mismatch during forward is unrelated to sinq quantized model save/load breakage." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 output corruption and return_tensors being ignored are different bugs." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer BPE marker handling and Qwen3.5 flash-attention generation are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Documentation/example mistakes are not the same as the processor.tokenizer regression in custom model code." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and DeepSpeed gradient inflation are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Embedding dtype drift and abnormal gradient growth under ZeRO-3 are different problems." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "A dtype change in evaluate is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model initialization breakage and lm_head.weight serialization inconsistency affect different code paths." + }, + { + "left": "issue:46209", + "right": "issue:46710", + "accept": false, + "reason": "FP8 inference on B200 and DeepSeek-R1 accuracy regression after upgrade are different issues." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but they hit different models with different special-token/BPE handling bugs." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig save/load token ID loss and MLukeTokenizer AttributeError are unrelated." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "A MacOS bus error in CLIP and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 hidden-states behavior and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A janus model test failure and a PEFT version bump are not the same bug." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Qwen2-57B tensor-parallel OOM and PEFT version pinning are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different tokenization rules; not the same underlying bug." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "StableLmConfig pad_token_idx removal and init_empty_weights argument mismatch are different 5.x breakages." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Specific model loading failure and HF_HUB_OFFLINE regression in _patch_mistral_regex are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing Qwen2_5_VLForConditionalGeneration and missing Float8 storage object are unrelated errors." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage lookup failure and flash_attn KeyError on Python 3.13 are different runtime issues." + }, + { + "left": "issue:46821", + "right": "issue:46832", + "accept": false, + "reason": "TimesFM window_size AttributeError and RT-DETR forward TypeError affect different models and code paths." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and incorrect visual encoder keys on save_pretrained are distinct regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 28, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:34567", + "issue:38175", + "issue:42490", + "issue:43257", + "issue:43299", + "issue:43334", + "issue:43493", + "issue:43504", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43696", + "issue:43701", + "issue:43844", + "issue:43873", + "issue:43937", + "issue:43994", + "issue:44112", + "issue:44464", + "issue:44493", + "issue:44849", + "issue:44869", + "issue:44964", + "issue:45072", + "issue:45084", + "issue:45200", + "issue:45229", + "issue:45245", + "issue:45520", + "issue:45584", + "issue:45588", + "issue:45759", + "issue:45823", + "issue:46050", + "issue:46285", + "issue:46399", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46832", + "issue:46846", + "issue:47145" + ], + "soft_pairs": [ + "issue:43334|issue:45759", + "issue:43504|issue:43606", + "issue:43334|issue:43572", + "issue:45245|issue:46710", + "issue:44493|issue:46285", + "issue:43257|issue:43299", + "issue:43504|issue:43653", + "issue:43606|issue:44112", + "issue:44849|issue:47145", + "issue:45072|issue:46832", + "issue:46399|issue:46846", + "issue:44464|issue:45584", + "issue:43844|issue:44964", + "issue:43873|issue:46050", + "issue:43994|issue:44493", + "issue:29159|issue:32090", + "issue:43937|issue:46701", + "issue:44869|issue:46752", + "issue:45200|issue:45229", + "issue:45229|issue:45823", + "issue:42490|issue:43696", + "issue:45084|issue:45520", + "issue:45588|issue:46050", + "issue:28218|issue:34567", + "issue:29942|issue:30333", + "issue:15354|issue:38175", + "issue:43493|issue:43701" + ], + "prepared_review_unit_hash": "b83ee30c7b4bbea5ddba94790b69c48940db45ca1282c3f25a48fd6b2be08b36", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35867, + "estimated_input_tokens": 8967, + "estimated_eval_tokens": 18190 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous rather than a duplicate set. The soft edges are mostly title-level similarities across different models or subsystems, but they describe different failure modes and fixes, so all soft pairs are rejected.", + "confidence": 0.96, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated bugs in tokenizers, Qwen loading, CI, quantization, generation, and config handling.", + "canonical_pr_reason": "No PRs are present in this cluster.", + "best_issue_reason": "No single issue is a good global representative because the cluster does not share one underlying bug or change.", + "best_pr_reason": "No PRs are present in this cluster.", + "soft_edge_verdicts": [ + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Both involve Qwen multimodal loading, but one is a missing pad_token_id attribute in Qwen3-VL and the other is from_config not unwrapping text_config for Qwen 3.5/3.6; different root causes." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "BEIT legacy-field loading failure and a Bark CPU-offload device mismatch are unrelated code paths and fixes." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both are config-attribute regressions after version updates, but they affect different models and different missing fields." + }, + { + "left": "issue:45245", + "right": "issue:46710", + "accept": false, + "reason": "A category-count runtime limit and a post-upgrade accuracy regression are not the same bug or fix." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id keys and CLIP text-model structure inconsistency are different API/shape issues." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Both mention Qwen MOE loading, but one is accelerate+deepspeed weight conversion and the other is a dev-version loading breakage for Qwen3VL; not the same concrete failure." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Legacy-field model loading and BigBirdTokenizer special-token registration are unrelated bugs." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Both mention device-related CI failures, but they are for different models and different mismatch mechanisms." + }, + { + "left": "issue:44849", + "right": "issue:47145", + "accept": false, + "reason": "Qwen3.5 hidden-states output bug and FP8 compressed-tensors weight_scale omission are different inference/serialization problems." + }, + { + "left": "issue:45072", + "right": "issue:46832", + "accept": false, + "reason": "A dtype mismatch in inference and an RT-DETR feature-level mismatch in forward are unrelated code-path issues." + }, + { + "left": "issue:46399", + "right": "issue:46846", + "accept": false, + "reason": "Missing decorator for apply_rotary_pos_emb registration and missing linear_attn entries in tp plan are different Qwen3.5 failures." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation with compiled forward and Whisper empty-transcription failure after align_special_tokens are unrelated." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "Gradient growth with ZeRO-3 and a Phi-4 multimodal load failure are different bugs in different subsystems." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Quantization offloading not working and sinq save/load corruption are related broadly, but they are different concrete failure modes and fixes." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 output mismatch and unexpected position-id keys are not the same underlying issue." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency and _gpu_broadcast_one rejecting NoneType are unrelated bugs." + }, + { + "left": "issue:43937", + "right": "issue:46701", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 and a DiffusionGemma pipeline failure are different problems." + }, + { + "left": "issue:44869", + "right": "issue:46752", + "accept": false, + "reason": "Whisper timestamp decoding crash and apply_chat_template on empty list are distinct edge cases in different APIs." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Defaulting mm_token_type_ids and a multi-GPU CUDA OOM are not the same bug." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Both concern Gemma4 inference, but one is OOM and the other is a device mismatch in PLE forward; different fixes." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Import warning for trailing slash paths and GPT-oss-20b CUDA OOM are unrelated." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Compilation of non-template nodes and a Python 3.13 flash_attn import KeyError are different issues." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention integration crashing on s_aux=None and quantized sinq save/load breakage are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer spacing after added tokens and TrainerState token-count tracking are different tokenizer/training-state bugs." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Failing Flash Attention 2 tests and MLFlow not marking failed jobs as finished are unrelated CI/tracking issues." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are unrelated model/runtime bugs." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 HF vs JAX discrepancy and resume_from_checkpoint key mismatch do not share the same code-path or 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": 27, + "review_unit_index": 29, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28218", + "issue:29870", + "issue:29942", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:43065", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43824", + "issue:43844", + "issue:43873", + "issue:43957", + "issue:44186", + "issue:44423", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44734", + "issue:44977", + "issue:44987", + "issue:45072", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45381", + "issue:45406", + "issue:45446", + "issue:45520", + "issue:45588", + "issue:45759", + "issue:45823", + "issue:46121", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46369", + "issue:46701", + "issue:46710", + "issue:46821" + ], + "soft_pairs": [ + "issue:44554|issue:45446", + "issue:43065|issue:46369", + "issue:41720|issue:46710", + "issue:44977|issue:46701", + "issue:44423|issue:44734", + "issue:37428|issue:43957", + "issue:43295|issue:43450", + "issue:44977|issue:46710", + "issue:45072|issue:45198", + "issue:45588|issue:45759", + "issue:45406|issue:46121", + "issue:43540|issue:45200", + "issue:43572|issue:46821", + "issue:43295|issue:43873", + "issue:44186|issue:46192", + "issue:43295|issue:44610", + "issue:45823|issue:46192", + "issue:43550|issue:43844", + "issue:43572|issue:44987", + "issue:43450|issue:43575", + "issue:29870|issue:29942", + "issue:45520|issue:46178", + "issue:37428|issue:43824", + "issue:45200|issue:45823", + "issue:42490|issue:43493", + "issue:39692|issue:46192", + "issue:43295|issue:43493", + "issue:42915|issue:43824", + "issue:44521|issue:45290", + "issue:28218|issue:35141", + "issue:45381|issue:46209" + ], + "prepared_review_unit_hash": "5334fc1cf7c254a03a5e32fac7f5689bc47acd7aba334f2d9eef5679251ca517", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37152, + "estimated_input_tokens": 9288, + "estimated_eval_tokens": 18832 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "None of the proposed soft pairs look like true duplicates; they span different models, subsystems, and failure modes with only superficial topic overlap.", + "confidence": 0.98, + "canonical_issue_reason": "No single issue is a clear canonical duplicate target here; the items are mostly unrelated bugs rather than one underlying problem.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue stands out as a global representative because the cluster does not converge on one concrete bug or code path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "Different problems: MPS attention correctness vs a PyTorch version-check/import issue in flex_attention." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Both mention Sam3, but one is about a dummy Conv2d in the pixel decoder and the other is a memory overflow in sam3-litetext." + }, + { + "left": "issue:41720", + "right": "issue:46710", + "accept": false, + "reason": "One is a CUDA assert during auto device mapping; the other is an accuracy regression after a version upgrade." + }, + { + "left": "issue:44977", + "right": "issue:46701", + "accept": false, + "reason": "Different models and failures: Qwen3.5 flash-attention generation vs a DiffusionGemma image-text pipeline issue." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both are serve-related, but they hit different code paths: continuous batching multimodal crash vs KV-cache continuation indexing." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "An import error for a flash-attention helper is not the same as meta-device model loading breaking some models." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Different regressions: processor/tokenizer access in custom code vs batched video processor output shape." + }, + { + "left": "issue:44977", + "right": "issue:46710", + "accept": false, + "reason": "Unrelated: flash-attention generation bug in Qwen3.5 vs an accuracy regression in DeepSeek." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "Different subsystems and symptoms: dtype mismatch in inference vs save_pretrained/tokenization failure." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "Different bugs: flash_attention integration crash on s_aux=None vs not unwrapping text_config for composite multimodal configs." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4 serve processor attribute error is unrelated to rope-parameter validation of JSON-loaded lists." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Video processing in Qwen3OmniMoe is unrelated to Gemma4 mm_token_type_ids defaults." + }, + { + "left": "issue:43572", + "right": "issue:46821", + "accept": false, + "reason": "StableLm pad_token_idx config regression and TimesFM window_size attribute error are distinct." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression is unrelated to quantization offloading behavior." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "Tokenizer crash on NER/padding is not the same as return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Different issues: processor.tokenizer regression vs OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device-mismatch during forward is unrelated to return_tensors handling." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "One is a torch.compile/SDPA bug; the other is a training gradient issue with ZeRO-3." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "StableLm config padding regression and a physical-intelligence loading failure are separate bugs." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Batched video shape bug and tensor-parallel OOM are unrelated failures." + }, + { + "left": "issue:29870", + "right": "issue:29942", + "accept": false, + "reason": "BART decoder_attention_mask UnboundLocalError is not the same as Flash Attention 2 test failures." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Both involve import/backends mapping, but they concern different backends and different error conditions." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Different ImportErrors for different missing symbols/modules." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Same model family, but one is a missing default token-type tensor and the other is a forward device placement bug." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Local-path import warning is unrelated to the SigLIP2 implementation discrepancy." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "Documentation/example errors for SigLIP2 are not the same as return_tensors being ignored." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "Processor/tokenizer regression and SigLIP2 parity discrepancy are different bugs." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "FineGrainedFP8Config failure in Qwen3Moe is unrelated to the Qwen2.5-VL import error." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both involve apply_chat_template, but one is wrong assistant masks for multimodal inputs and the other is a crash on tool-call messages with no content." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer spacing after added tokens and embedding reinitialization after resize are different bugs." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Qwen2.5-VL video position IDs 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": 27, + "review_unit_index": 30, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:37428", + "issue:41720", + "issue:43262", + "issue:43278", + "issue:43334", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43450", + "issue:43452", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43819", + "issue:43824", + "issue:43827", + "issue:43856", + "issue:43901", + "issue:43950", + "issue:43975", + "issue:44112", + "issue:44442", + "issue:44466", + "issue:44561", + "issue:44610", + "issue:44737", + "issue:44792", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45325", + "issue:45412", + "issue:45563", + "issue:45593", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46512", + "issue:46894", + "issue:46899", + "issue:46932" + ], + "soft_pairs": [ + "issue:44466|issue:45127", + "issue:44737|issue:46894", + "issue:43278|issue:43856", + "issue:43653|issue:44112", + "issue:43824|issue:43975", + "issue:43452|issue:46512", + "issue:43262|issue:43450", + "issue:43950|issue:46285", + "issue:43334|issue:43525", + "issue:44442|issue:45198", + "issue:43526|issue:46192", + "issue:45200|issue:46899", + "issue:45200|issue:46932", + "issue:44792|issue:45563", + "issue:44898|issue:45412", + "issue:43819|issue:44112", + "issue:43335|issue:43550", + "issue:45200|issue:45325", + "issue:44561|issue:46234", + "issue:41720|issue:46209", + "issue:44610|issue:44857", + "issue:43819|issue:45083", + "issue:43827|issue:43901", + "issue:43606|issue:45083", + "issue:44857|issue:45593", + "issue:43388|issue:43526", + "issue:37428|issue:43425", + "issue:43335|issue:43653" + ], + "prepared_review_unit_hash": "cda73577c37f843c8d9b96bfa411b06e9b3519c4a421c14c28ea72e1c25aec14", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35952, + "estimated_input_tokens": 8988, + "estimated_eval_tokens": 18232 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: the items span unrelated regressions in model configs, tokenizers, processors, training, docs, and CI. None of the soft pairs describe the same concrete bug or change, so all soft edges are rejected and there is no clear canonical issue/PR for the cluster.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Both concern tied-weight behavior, but one is about device-dependent serialization of lm_head.weight and the other about LoRA merge collapse with extended vocab; different failure modes and fixes." + }, + { + "left": "issue:44737", + "right": "issue:46894", + "accept": false, + "reason": "XLNet CPU arange placement is unrelated to CohereASR's training-loss double-shift bug; different models, code paths, and symptoms." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "BF16\u2192FP32 dtype drift in evaluate is not the same as Qwen3 MoE memory inefficiency during training." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "A tokenizer special-token registration bug is unrelated to a GraniteSpeech CI device-override test regression." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "Missing Qwen2_5_VL import/export is a namespace/API break; incorrect detokenization in deepseek-coder is a separate tokenizer behavior bug." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file/from_pretrained loading issues are unrelated to from_config(dtype=...) leaving config.dtype stale." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate defaults and batched video output shapes are different processor bugs." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Non-persistent buffer corruption during from_pretrained is unrelated to CLIP text-model structural inconsistency." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention pad_token_id, but they are separate missing-attribute bugs in different model configs (Qwen3-VL vs Llama4)." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer loading failure is unrelated to Wav2Vec2 save_pretrained/tokenization failure." + }, + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and text_kwargs/return_tensors handling are distinct preprocessing issues." + }, + { + "left": "issue:45200", + "right": "issue:46899", + "accept": false, + "reason": "Defaulting mm_token_type_ids to zeros for text-only fine-tuning is unrelated to unsafe dtype casting in quantized multimodal embedders." + }, + { + "left": "issue:45200", + "right": "issue:46932", + "accept": false, + "reason": "These are different Gemma4 bugs: one is text-only mm_token_type_ids defaults, the other is quantized vision/audio embedder casting." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "A failed janus image-generation test is not the same as a stale paged-generate warning for num_return_sequences." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver non-default resolution handling and RT-DETR memory release are unrelated model issues." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC.from_latents mismatch with missing STE is unrelated to a GraniteSpeech CI test failure." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and Bamba SDPA torch.compile failure affect different models and code paths." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Gemma4 text-only mm_token_type_ids defaults are unrelated to Qwen2.5-VL rope index scaling for still images." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "These are separate import_utils API removals/additions affecting different helpers and remote-code models." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3 auto device-mapping cuda asserts on A800 and FP8 inference failure on B200 are different hardware/model issues." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "OmDet-Turbo processor size mismatch and LwDetrImageLoss AMP crash are unrelated vision bugs." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "DAC latent reconstruction mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs regressions, but they affect different pipeline pages and different deprecations; not the same change." + }, + { + "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." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "AMP crash in LwDetrImageLoss is not the same underlying issue as D-FINE omitting auxiliary losses when denoising is off." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Mask2Former gather_for_metrics label truncation is unrelated to BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "A missing flash-attention helper import and broad Torch 2.10 incompatibility are different compatibility problems." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Sparse-layer misconfiguration in SwitchTransformers and BigBirdTokenizer special-token registration are unrelated 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": 27, + "review_unit_index": 31, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43819", + "issue:43825", + "issue:43854", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44448", + "issue:44466", + "issue:44493", + "issue:44514", + "issue:44521", + "issue:44743", + "issue:44855", + "issue:44857", + "issue:45083", + "issue:45290", + "issue:45412", + "issue:45520", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:45859", + "issue:46170", + "issue:46285", + "issue:46326", + "issue:46424", + "issue:46512", + "issue:46821", + "issue:46858" + ], + "soft_pairs": [ + "issue:44206|issue:45584", + "issue:46512|issue:46821", + "issue:45759|issue:46285", + "issue:46424|issue:46858", + "issue:29942|issue:32090", + "issue:38175|issue:45412", + "issue:28282|issue:33357", + "issue:43335|issue:43504", + "issue:43611|issue:44466", + "issue:44855|issue:45083", + "issue:43550|issue:44112", + "issue:45520|issue:46170", + "issue:37428|issue:44351", + "issue:43335|issue:45083", + "issue:44448|issue:44743", + "issue:43335|issue:44112", + "issue:43335|issue:45593", + "issue:43278|issue:43825", + "issue:34689|issue:35141", + "issue:42490|issue:43854", + "issue:45290|issue:46326", + "issue:45588|issue:45859", + "issue:29127|issue:32090", + "issue:44514|issue:44521", + "issue:39692|issue:43526", + "issue:43606|issue:44857", + "issue:43819|issue:44493" + ], + "prepared_review_unit_hash": "32778769ea29b598b25fb410eb03766298a700194397e92afeea9901a23a1b36", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36239, + "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 broad and mostly heterogeneous: many distinct Transformers bugs around loading/configuration, chat templates, cache behavior, and backend-specific failures. There is no clear duplicate pair among the soft candidates, so the cluster should be treated as noisy rather than a single underlying defect.", + "confidence": 0.34, + "canonical_issue_reason": "issue:43611 is the broadest, most cross-cutting regression in the set (model loading with base_model_prefix in v5) and is a better representative than the many narrowly scoped model-specific or backend-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single issue representative because it describes a general loading regression affecting multiple models, which is closer to the cluster\u2019s load/config-heavy center of mass than the very specific edge cases elsewhere.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Both are runtime crashes, but one is a feature-extractor argument issue and the other is an empty-transcription generation bug; different code paths and fixes." + }, + { + "left": "issue:46512", + "right": "issue:46821", + "accept": false, + "reason": "Stale dtype metadata in from_config and a TimesFM window_size AttributeError are unrelated bugs despite both being configuration-facing." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "One is about AutoModelForCausalLM unwrapping text_config for composite multimodal configs; the other is a CLIP class structure inconsistency. Not the same defect." + }, + { + "left": "issue:46424", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache mechanics, but one is about max_cache_length handling and the other is a device-tensor seq_length arithmetic bug in static cache; distinct failures." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention test failures and a Trainer _gpu_broadcast_one NoneType error are unrelated issues in different subsystems." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Unexpected zero probabilities and an RT-DETR memory leak are completely different model/runtime problems." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "Missing PyTorch ImportError and a MacOS bus error with a community CLIP model are not the same bug." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and BEiT legacy-field loading are different model-specific configuration bugs." + }, + { + "left": "issue:43611", + "right": "issue:44466", + "accept": false, + "reason": "Model-loading failure via base_model_prefix is not the same as inconsistent lm_head.weight serialization depending on device." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 IndentationError in DebertaV2Model and a qwen3_omni_moe helper output-length bug are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "torch.compile/SDPA failure for Bamba-9B-v2 and a stale-device override CI failure are different issues." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "A flash_attn import key error on Python 3.13 and a Windows ROCm torch.distributed import/type error are separate platform-specific problems." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Cannot import _flash_supports_window_size and cannot import HybridCache are both import errors, but they stem from different symbols and modules." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "A sparse-layer creation bug and a helper function returning wrong feature lengths are unrelated." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Pegasus output differences across versions and qwen3 recurrent state resets are different model behaviors with different fixes." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformersConfig sparse-layer initialization and GraniteSpeech stale device override are not the same code-path problem." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Sparse encoder-layer creation and D-FINE auxiliary-loss handling when denoising is off are unrelated model-specific bugs." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Embedding dtype staleness during evaluate and a pipeline task-support warning bug do not share the same underlying cause." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Llama 3.2 Vision model loading regression and token-embedding resize/post_init reinitialization are different issues." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "Trailing-slash local model import warning and GLM-4.7-Flash test-loading failures are distinct loading problems." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both concern apply_chat_template, but one crashes on assistant tool-call messages with no content while the other is a mistral_common continue_final_message bug; not the same defect." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "A flash_attention integration crash on s_aux=None and a Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are different backend issues." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and Trainer GPU broadcast NoneType support are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both are Qwen2_5_VL apply_chat_template problems, but one is a batching/padding crash and the other is all-zero assistant masks; similar area, different bug." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "SigLIP2 docs-example errors and BEiT fast image-processor reduce_labels behavior are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU-offload device mismatch in bark-small and float16 AMP crash in LwDetrImageLoss are different hardware/backend bugs." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC.from_latents/STE mismatch and unexpected position-id keys are unrelated model internals." + } + ] + }, + "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": 27, + "review_unit_index": 32, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:26413", + "issue:27957", + "issue:28218", + "issue:29870", + "issue:29942", + "issue:31501", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:42175", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43550", + "issue:43575", + "issue:43701", + "issue:43720", + "issue:44112", + "issue:44206", + "issue:44448", + "issue:44464", + "issue:44493", + "issue:44617", + "issue:44779", + "issue:44857", + "issue:45072", + "issue:45083", + "issue:45127", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", + "issue:45593", + "issue:45823", + "issue:46234", + "issue:46424", + "issue:46431", + "issue:46566", + "issue:46697", + "issue:47000" + ], + "soft_pairs": [ + "issue:46431|issue:46566", + "issue:28218|issue:32090", + "issue:45310|issue:45478", + "issue:43720|issue:44112", + "issue:39692|issue:44493", + "issue:44448|issue:44779", + "issue:43701|issue:45083", + "issue:44206|issue:45823", + "issue:44617|issue:46697", + "issue:26413|issue:31501", + "issue:45127|issue:45245", + "issue:45459|issue:46234", + "issue:32090|issue:33357", + "issue:42490|issue:43575", + "issue:44857|issue:45072", + "issue:43550|issue:45083", + "issue:45593|issue:47000", + "issue:44464|issue:46424", + "issue:27957|issue:34689", + "issue:38175|issue:44493", + "issue:29942|issue:35141", + "issue:29870|issue:32090", + "issue:10000|issue:15354", + "issue:43278|issue:43329", + "issue:45290|issue:45381", + "issue:37428|issue:42175" + ], + "prepared_review_unit_hash": "e054b112e26cfccf0b6b83a7f7d42abc39450234d2a6b3ce3d9e836fbbc4cb4b", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35566, + "estimated_input_tokens": 8892, + "estimated_eval_tokens": 18040 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are unrelated and should stay separate. The only clear duplicate is the two Qwen3.5 Moe from_pretrained reports: issue:45310 and issue:45478.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45310 is the earlier report in the same Qwen3.5 Moe from_pretrained error family, with a matching title to issue:45478.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the strongest canonical issue candidate because it is the earlier report and clearly names the same Qwen3.5 Moe from_pretrained failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46431", + "right": "issue:46566", + "accept": false, + "reason": "Both touch model-loading behavior, but one is about a missing chat template error and the other is a device_map/meta-device placement bug; different failures and code paths." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer spacing after added tokens is unrelated to a Trainer NoneType broadcast error." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same model family and same from_pretrained error theme, with matching titles; issue:45310 is the earlier report." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "Packed-weight loading in accelerate and a GraniteSpeech CI device-override test are different bugs." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 docs/model-processor mismatch and unexpected position-id keys are not the same underlying issue." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Pegasus v4/v5 output divergence and Deepseek tokenizer regression are different code paths." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "resume_from_checkpoint key mismatch is unrelated to qwen3_omni_moe feature-length behavior." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Unsupported center argument in LasrFeatureExtractor and Gemma4 device mismatch are distinct bugs." + }, + { + "left": "issue:44617", + "right": "issue:46697", + "accept": false, + "reason": "Sam3Video CUDA OOM and Swin2SR over-padding are unrelated." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Both relate to training startup, but the concrete failure modes are different." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge output collapse and category-count overflow are separate problems." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Tokenizer protobuf error handling and a missing import_utils backend symbol are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Trainer broadcast NoneType and MacOS CLIP bus error are different runtime failures." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Local-path import warning and TP OOM are not the same bug." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "AMP/CUDA loss crash and bfloat16 dtype mismatch in CI are different issues." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile + SDPA failure and feature-length helper behavior are unrelated." + }, + { + "left": "issue:45593", + "right": "issue:47000", + "accept": false, + "reason": "Auxiliary-loss behavior and infeasible cost-matrix crash are different detection bugs." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation inconsistencies and cache max-length handling are separate generation/cache issues." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta and Llama 3.2 Vision loading regression are not the same issue." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and unexpected key/position-id issues do not match." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Failing Flash Attention 2 tests and output embedding reinitialization after resizing are unrelated." + }, + { + "left": "issue:29870", + "right": "issue:32090", + "accept": false, + "reason": "BART generate decoder_attention_mask assignment and Trainer broadcast error are different bugs." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "German DistilBertModel issue and torch.jit.script GeneratorExp unsupported are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift during evaluate and undefined video-token helper variables are different issues." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template tool-call crash and qwen2.5-vl video position-id bug are separate code paths." + }, + { + "left": "issue:37428", + "right": "issue:42175", + "accept": false, + "reason": "Missing _flash_supports_window_size import and missing TensorFlow in the torch extra are unrelated packaging/import problems." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The duplicate call on issue:45310 / issue:45478 is plausible, but the writeup overstates the evidence. The packet only shows identical titles, close dates, and similar body lengths; it does not justify \u201cnear-verbatim\u201d or \u201cmore complete.\u201d Please keep the canonical rationale more conservative, e.g. earlier report with matching title and same model/error family." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 33, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:25251", + "issue:29127", + "issue:29870", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34567", + "issue:39692", + "issue:40444", + "issue:41720", + "issue:42490", + "issue:43278", + "issue:43381", + "issue:43526", + "issue:43575", + "issue:43611", + "issue:43701", + "issue:43873", + "issue:43927", + "issue:43976", + "issue:44186", + "issue:44360", + "issue:44361", + "issue:44448", + "issue:44466", + "issue:44561", + "issue:44568", + "issue:44779", + "issue:44855", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45229", + "issue:45290", + "issue:45440", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45698", + "issue:45803", + "issue:45823", + "issue:45920", + "issue:46121", + "issue:46192", + "issue:46399", + "issue:46421", + "issue:46655", + "issue:46858" + ], + "soft_pairs": [ + "issue:44448|issue:44977", + "issue:29870|issue:33290", + "issue:44855|issue:46399", + "issue:44360|issue:44561", + "issue:43278|issue:43381", + "issue:40444|issue:43526", + "issue:32090|issue:33290", + "issue:45920|issue:46192", + "issue:45290|issue:46655", + "issue:43927|issue:45698", + "issue:44977|issue:45229", + "issue:42490|issue:43873", + "issue:46421|issue:46858", + "issue:43873|issue:45229", + "issue:25251|issue:32090", + "issue:44448|issue:44466", + "issue:43381|issue:43701", + "issue:43976|issue:45440", + "issue:44871|issue:44977", + "issue:29127|issue:34567", + "issue:33357|issue:39692", + "issue:43278|issue:43575", + "issue:44561|issue:44779", + "issue:44568|issue:45491", + "issue:44361|issue:45198", + "issue:45588|issue:45823", + "issue:41720|issue:45229", + "issue:44186|issue:45198", + "issue:45198|issue:45803", + "issue:44977|issue:45563", + "issue:43611|issue:46121" + ], + "prepared_review_unit_hash": "471fb51b7fa2756b7b42416a8a8a41fe9ea45cee50c7a76786a91b1279051c50", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37315, + "estimated_input_tokens": 9329, + "estimated_eval_tokens": 18914 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated standalone issues across different models, tokenizer paths, and runtime subsystems. The only broadly representative issue is the v5 AutoTokenizer regression affecting multiple model families.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45920 has the broadest scope and strongest evidence of a shared underlying bug: a v5 regression affecting token IDs across multiple model families, with inbound references and clearer generality than the other issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best representative issue because it covers a concrete, widely-impacting tokenizer regression rather than a model-specific edge case.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and different failure modes: Pegasus output drift vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:29870", + "right": "issue:33290", + "accept": false, + "reason": "Unrelated bugs: BART generate() unbound local error vs deepspeed/Adafactor OOM." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Both are import/runtime errors, but on different models and different code paths; not the same underlying bug." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "Completely different areas: DSA indexer discussion vs v5 trust_remote_code FX helper removal." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Both touch eval mode, but one is a dtype regression and the other is a gradient-checkpointing restriction." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Different vision preprocessing bugs in different components/models." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Both involve training infrastructure, but one is a NoneType broadcast error and the other is an optimizer memory issue." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Both are tokenizer-related, but one is a broad wrong-token-ID regression and the other is a return_tensors handling bug." + }, + { + "left": "issue:45290", + "right": "issue:46655", + "accept": false, + "reason": "Different subsystems and symptoms: chat template/tool-call crash vs video processor padding calculation." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load behavior, but one is config token IDs and the other is wrong custom module loading." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Unrelated model-specific generation bug vs multi-GPU OOM." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Path parsing warning is unrelated to quantization/offloading behavior." + }, + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache/compile, but one is spurious recompilation and the other is a tensor-type arithmetic bug in prefill." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Different resource issues: quantization/offloading mismatch vs Gemma4 multi-GPU OOM." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to trainer GPU broadcast NoneType error." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Both are v5-related, but one is generation output divergence and the other is serialization of lm_head.weight." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Different training-state bugs: eval-mode checkpointing vs resume checkpoint key mismatch." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "One is a Python version compatibility regression, the other a model implementation divergence." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and different issues: eos_token_id mismatch vs flash-attention generation failure." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error-message clarity is unrelated to TrainerState token counter updates." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "Both are vision-model related, but one is a MacOS bus error and the other is a documentation-example mismatch/quantization failure." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Different problems: eval dtype drift vs tensor-parallel load OOM." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Both are v5 regressions, but one breaks remote-code model loading and the other is tokenizer output correctness for DeepSeek." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer special-token insertion bug is unrelated to Gemma3 NaN embeddings on mixed-length batches." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer/processor bugs in different model families and code paths." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Both are forward/runtime model issues, but they affect different mechanisms and are not the same concrete bug." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Different hardware/resource failures: cudaErrorAssert with auto device mapping vs multi-GPU OOM." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Different model/tokenizer failure modes; no evidence of a shared underlying bug." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Tokenization/save-pretrained failure is unrelated to a broad exception-handling bug in FuyuBatchFeature." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Generation crash vs a stale warning; not the same issue." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "Base-model-prefix loading regression is unrelated to rope-parameter validation on JSON-loaded lists." + } + ] + }, + "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": 27, + "review_unit_index": 34, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:29870", + "issue:29942", + "issue:30064", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43066", + "issue:43381", + "issue:43412", + "issue:43421", + "issue:43526", + "issue:43575", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43994", + "issue:44206", + "issue:44265", + "issue:44387", + "issue:44442", + "issue:44462", + "issue:44479", + "issue:44743", + "issue:45072", + "issue:45198", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45593", + "issue:45712", + "issue:45834", + "issue:46369", + "issue:46693", + "issue:46701", + "issue:47217" + ], + "soft_pairs": [ + "issue:29870|issue:34689", + "issue:29942|issue:30064", + "issue:43653|issue:44743", + "issue:44206|issue:44479", + "issue:42915|issue:42994", + "issue:43756|issue:45440", + "issue:43653|issue:44442", + "issue:44265|issue:45198", + "issue:43575|issue:43701", + "issue:29127|issue:33357", + "issue:43065|issue:45412", + "issue:43065|issue:43994", + "issue:45491|issue:46693", + "issue:43412|issue:45588", + "issue:42490|issue:43381", + "issue:45372|issue:45588", + "issue:29127|issue:35141", + "issue:45245|issue:45563", + "issue:45459|issue:45712", + "issue:45440|issue:45593", + "issue:43526|issue:47217", + "issue:43994|issue:45412", + "issue:43575|issue:44387", + "issue:43421|issue:45491", + "issue:45406|issue:46701", + "issue:45372|issue:45406", + "issue:43066|issue:44462", + "issue:45412|issue:46369", + "issue:44442|issue:45072", + "issue:45372|issue:45459", + "issue:27957|issue:33290", + "issue:43526|issue:45834", + "issue:27957|issue:29127", + "issue:43526|issue:46693", + "issue:33290|issue:34689" + ], + "prepared_review_unit_hash": "2a36c6592ef7a953f048ef39f79c640d1feece4e32a5ee56e09d5c2e225bef27", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 38136, + "estimated_input_tokens": 9534, + "estimated_eval_tokens": 19324 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated bugs across different models, tokenizers, processors, export paths, and memory behaviors. The soft-similarity pairs share broad subsystem words, but not the same concrete defect, so none should be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "No strong canonical issue: the set is heterogeneous and the suggested pairs do not indicate a single underlying bug family.", + "canonical_pr_reason": "No pull requests are present in this cluster.", + "best_issue_reason": "issue:44462 is one of the clearest, most concrete reports here, with a specific reproducible loading bug and direct impact on AutoTokenizer behavior.", + "best_pr_reason": "No pull requests are present in this cluster.", + "soft_edge_verdicts": [ + { + "left": "issue:29870", + "right": "issue:34689", + "accept": false, + "reason": "Different failures in different areas: BART generate() decoder mask handling vs Llama 3.2 Vision model loading regression." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures and void segmentation-map processing are unrelated bugs." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Tokenizer special-token registration/empty decode is unrelated to Qwen recurrent state reset logic." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Feature extractor center-arg regression and video-input regression affect different code paths and models." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both involve quantization, but one is a FineGrainedFP8Config failure and the other is quantized saving; different concrete problems." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Different model-implementation mismatches: Smollm3 RoPE layer dropping vs DeepseekV3MoE divergence from remote behavior." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBird tokenizer special-token registration bug is not the same as FastSpeech2ConformerTokenizer loading failure." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export with torch_compilable_check is a different failure mode from Wav2Vec2 save_pretrained/tokenization issues." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "OOM during tensor-parallel model load is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and MacOS bus error on CLIP loading are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory release/leak behavior." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Sam3PixelDecoder dummy layer and SigLIP2 nonsensical outputs are different bugs in different models." + }, + { + "left": "issue:45491", + "right": "issue:46693", + "accept": false, + "reason": "Both mention long-context/FA2, but one is a Gemma3 NaN embedding bug and the other is a performance regression; not the same defect." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Gemma3n export guard/erfinv support and flash_attention s_aux=None AttributeError are distinct code-path failures." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-path import warning and gradient-checkpointing-in-eval-mode are unrelated." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 processor loading import error is unrelated to flash_attention sink-less model handling." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Error-message clarity in LayoutLMv3 is unrelated to post_init reinitializing resized embeddings." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-cardinality RuntimeError and stale paged-generate warning are different issues." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Tokenizer protobuf exception masking and leftover dummy PT classes are unrelated." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3MoE divergence and D-FINE auxiliary-loss behavior are different model bugs." + }, + { + "left": "issue:43526", + "right": "issue:47217", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to VL image-placeholder handling." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 wrong outputs and RT-DETR memory release are unrelated defects." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both can OOM, but tensor-parallel load OOM and int4 quantization reserved-memory regression have different root causes." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime special-token post-processor update bug is unrelated to Gemma3 mixed-length NaN embeddings." + }, + { + "left": "issue:45406", + "right": "issue:46701", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to DiffusionGemma image-text-to-text pipeline failure." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma 4 processor loading import error and Gemma4Processor missing _tokenizer are different failures." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring tokenizer.json from the repository." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "RT-DETR memory leak and Sam3-litetext memory overflow are different memory issues in different models." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer loading and bfloat16 dtype mismatches are unrelated." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma 4 import failure from mistral_common and protobuf decode-error masking affect different tokenizer/processor paths." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and Adafactor+DeepSpeed OOM are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and Kosmos2.5 long OCR index error are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta and LayoutLMv3 error messaging have no common underlying bug." + }, + { + "left": "issue:43526", + "right": "issue:46693", + "accept": false, + "reason": "Label reduction bug and FA2 long-context performance regression are unrelated." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "DeepSpeed Adafactor OOM and Llama 3.2 Vision loading breakage are different 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": 27, + "review_unit_index": 35, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:27957", + "issue:29870", + "issue:33290", + "issue:35141", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43316", + "issue:43334", + "issue:43381", + "issue:43493", + "issue:43525", + "issue:43531", + "issue:43646", + "issue:43720", + "issue:43761", + "issue:43824", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44077", + "issue:44206", + "issue:44279", + "issue:44442", + "issue:44479", + "issue:44514", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44871", + "issue:44898", + "issue:44977", + "issue:45003", + "issue:45245", + "issue:45250", + "issue:45381", + "issue:45440", + "issue:45588", + "issue:45859", + "issue:46121", + "issue:46234", + "issue:46369", + "issue:46399", + "issue:46640", + "issue:46821" + ], + "soft_pairs": [ + "issue:43824|issue:44279", + "issue:33290|issue:35141", + "issue:44857|issue:45588", + "issue:42915|issue:43931", + "issue:43994|issue:46369", + "issue:43334|issue:43531", + "issue:45250|issue:46640", + "issue:43381|issue:43873", + "issue:45588|issue:46821", + "issue:46121|issue:46399", + "issue:44514|issue:45381", + "issue:43646|issue:44077", + "issue:44977|issue:45245", + "issue:43761|issue:45859", + "issue:44743|issue:44811", + "issue:44442|issue:44898", + "issue:43262|issue:44206", + "issue:42490|issue:44743", + "issue:43720|issue:44442", + "issue:27957|issue:29870", + "issue:42175|issue:46234", + "issue:44077|issue:44479", + "issue:43316|issue:43525", + "issue:45003|issue:45440", + "issue:43493|issue:43827", + "issue:43873|issue:44871" + ], + "prepared_review_unit_hash": "016d4ef93c18ed0e53b698668ddf457eeb8b7b110b0c1b2dc075698bf883e5f2", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34814, + "estimated_input_tokens": 8704, + "estimated_eval_tokens": 17664 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is largely heterogeneous: the soft pairs look like distinct bugs or regressions in different code paths, so I would reject every soft edge. There are no pull requests in the cluster. If one issue has to serve as a representative, 43873 is the most central/active item, but it is not a true duplicate umbrella for the rest.", + "confidence": 0.41, + "canonical_issue_reason": "issue:43873 is the best representative only because it is active, has inbound references, and describes a broad runtime behavior regression; however, the cluster does not form one coherent duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the most suitable issue to anchor triage because it is open, discussion-heavy, and broader than the other items, which are mostly isolated model/config regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "One is a Qwen2.5-VL import error; the other is a vague dependency issue. No shared failing code path." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "AdaFactor/DeepSpeed OOM and embedding reinitialization in post_init are unrelated bugs." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "Different failures in different subsystems: LwDETR AMP/CUDA loss vs flash-attention sink-less model crash." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3MoE FP8 config failure and Qwen3-VL weight-shape mismatch are distinct model-loading problems." + }, + { + "left": "issue:43994", + "right": "issue:46369", + "accept": false, + "reason": "SigLIP2 incorrect outputs and SAM3 memory overflow are different models and different symptoms." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id in Qwen3-VL config is not the same as a sliding_window issue in Qwen3-MoE." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Both mention Flash Attention 2, but the titles are too generic to show the same concrete bug." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and quantization/offloading behavior are separate mechanisms." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "A flash-attention AttributeError and a TimesFM window_size AttributeError are unrelated failures." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "Rope-validation TypeError and Qwen3.5 apply_rotary_pos_emb registration failure are different code paths." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL regressions, but one is batched chat-template padding and the other is video vision_position_ids." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "A broad v5 custom-init break and a patchtsmixer post_init policy issue are not the same concrete change." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3.5 flash-attention generation failure and category-count overflow are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Qwen3.5 recurrent-state reset and Whisper batch_decode skip_special_tokens are different features." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver non-default resolution failure are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sampling-rate default and LaserFeatureExtractor center-arg crash are different audio bugs." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local-path import warning and Qwen3.5 recurrent-state reset do not share a code path." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight loading and tokenizer loading are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:29870", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and BART generate() decoder_attention_mask error are distinct issues." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Torch-only install backend packaging and missing import_utils symbol are not the same bug." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "PatchTSMixer post_init policy and Qwen multi-VL video regression are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Llama4Config missing pad_token_id are separate model config mismatches." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "sys.modules access safety and DeepSeekV3MoE divergence are different problems." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation mismatch and outdated docs referencing pipeline() are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization/offloading behavior and Gemma-3 eos_token_id configuration are different 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": 27, + "review_unit_index": 36, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:25251", + "issue:27957", + "issue:28282", + "issue:32090", + "issue:42617", + "issue:42907", + "issue:43232", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43643", + "issue:43701", + "issue:43742", + "issue:43749", + "issue:43761", + "issue:43825", + "issue:43931", + "issue:44062", + "issue:44079", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44496", + "issue:44521", + "issue:44855", + "issue:44871", + "issue:44913", + "issue:44987", + "issue:45137", + "issue:45216", + "issue:45325", + "issue:45584", + "issue:45698", + "issue:45759", + "issue:45803", + "issue:45907", + "issue:46178", + "issue:46506", + "issue:46697", + "issue:46829", + "issue:46932", + "issue:47217" + ], + "soft_pairs": [ + "issue:44062|issue:45907", + "issue:44855|issue:45137", + "issue:43643|issue:44913", + "issue:45325|issue:46697", + "issue:43316|issue:45759", + "issue:43749|issue:43931", + "issue:44987|issue:46178", + "issue:44190|issue:44496", + "issue:43931|issue:45216", + "issue:43701|issue:46506", + "issue:42617|issue:43575", + "issue:44442|issue:45803", + "issue:43550|issue:43701", + "issue:43232|issue:43388", + "issue:43575|issue:43742", + "issue:43761|issue:44079", + "issue:27957|issue:32090", + "issue:43761|issue:44188", + "issue:43606|issue:43701", + "issue:25251|issue:28282", + "issue:46829|issue:46932", + "issue:42907|issue:45698", + "issue:16998|issue:22355", + "issue:43295|issue:44871", + "issue:43295|issue:43825", + "issue:44521|issue:47217", + "issue:44871|issue:45584" + ], + "prepared_review_unit_hash": "44d4d6a34413369cb5529a4dfa3d30349466cc5748d969bff6f2217475d92dc2", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35292, + "estimated_input_tokens": 8823, + "estimated_eval_tokens": 17902 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items span many unrelated bugs and regressions, with no soft pair looking like the same underlying change strongly enough to merge. I\u2019d keep the cluster anchored on the most actionable, high-signal issue rather than treating any pair as duplicates.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43749 stands out as the clearest high-signal regression: it has the most discussion and inbound references, and the title points to a concrete broken feature rather than a vague environment or usage question.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the best active triage anchor because it is open, has multiple inbound references, and describes a specific reproducible multimodal API failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "Different code paths: tokenizer AddedToken keyword collision vs list_repo_files not forwarding revision; unrelated bugs." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 torch.jit parsing/import error vs DeepSpeed ZeRO3 deque underflow; no shared underlying defect." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "One is missing fields from AutoConfig with trust_remote_code, the other is GPTNeoX config reload losing rotary_pct; different config bugs." + }, + { + "left": "issue:45325", + "right": "issue:46697", + "accept": false, + "reason": "Qwen2.5-VL rope/temporal position scaling issue vs Swin2SR over-padding in image processor; different preprocessing logic." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy vs Qwen multimodal from_config text_config unwrapping; different model families and failure modes." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU RAM efficient loading regression is unrelated to a model-specific weight shape mismatch on load." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Loading a specific model on transformers>=5.1.0 vs tensorflow_text backend mapping error; no common bug." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Dataset loading in an example script vs AutoTokenizer rejecting an unrecognized model config; unrelated entry points." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Model weight shape mismatch on load vs corrupted save_pretrained checkpoint output; distinct load/save bugs." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "resume_from_checkpoint key mismatch vs CPU/CUDA output divergence in PPDocLayoutV3; unrelated." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "Failure to run a parallel script vs OOM when loading a large model with tensor parallelism; same broad area, different problem." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer loading failure vs bare except swallowing interrupts in Fuyu tensor conversion; unrelated." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile SDPA issue in Bamba vs checkpoint resume key mismatch; different subsystems." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs sync-gpus issue vs gather_for_metrics dropping tuple labels; different distributed-training bugs." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "OOM loading a large Qwen model vs key error loading MobileLLM-125M; not the same underlying issue." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression vs ModelOutput key assignment behavior when previous key is None; separate output-structure bugs." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta vs Trainer broadcast TypeError on None; unrelated." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "Missing hidden_states output vs attention kernel divergence under torch.compile; different code paths." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch in bark-small vs resume_from_checkpoint key mismatch; unrelated." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "pipeline top_k output shape regression vs missing PyTorch dependency error; unrelated." + }, + { + "left": "issue:46829", + "right": "issue:46932", + "accept": false, + "reason": "VideoPrism hidden_states misassignment vs Gemma4 quantized embedder dtype casting fix; distinct model internals." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Saving dequantized models correctly vs loading the wrong custom module after save_pretrained; both serialization-related but not the same bug." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length question for DeBERTa-V3 vs missing transformers.onnx module import; unrelated." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "processor.tokenizer/image-to-tokenizer regression vs Gemma-3 eos_token_id config mismatch; separate regression and config issues." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "Custom multimodal processor regression vs pipeline error message about translation support in v5; different user-facing failures." + }, + { + "left": "issue:44521", + "right": "issue:47217", + "accept": false, + "reason": "Both involve multimodal prompting, but one is all-zero assistant masks and the other is failure when an image placeholder appears in user text; not clearly the same defect." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma-3 eos_token_id inconsistency vs Whisper empty transcription after special-token alignment; unrelated special-token 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": 27, + "review_unit_index": 37, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:43065", + "issue:43316", + "issue:43352", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43504", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43572", + "issue:43646", + "issue:43716", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44464", + "issue:44496", + "issue:44514", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:44977", + "issue:45127", + "issue:45200", + "issue:45381", + "issue:45542", + "issue:45600", + "issue:45632", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46234", + "issue:46291", + "issue:46365", + "issue:46369", + "issue:46421", + "issue:46424", + "issue:46640", + "issue:46821", + "issue:46894", + "issue:47217" + ], + "soft_pairs": [ + "issue:46234|issue:46291", + "issue:43316|issue:43716", + "issue:43316|issue:46821", + "issue:44464|issue:46421", + "issue:43646|issue:44913", + "issue:44077|issue:44534", + "issue:43452|issue:44496", + "issue:44871|issue:46050", + "issue:45803|issue:46365", + "issue:44162|issue:44792", + "issue:43404|issue:43441", + "issue:44242|issue:46894", + "issue:43065|issue:44617", + "issue:45632|issue:46424", + "issue:43572|issue:46178", + "issue:43540|issue:44514", + "issue:44610|issue:46369", + "issue:45381|issue:47217", + "issue:45542|issue:46234", + "issue:44792|issue:45600", + "issue:43504|issue:45698", + "issue:45803|issue:46170", + "issue:44792|issue:44977", + "issue:43352|issue:46640", + "issue:45127|issue:45200", + "issue:43525|issue:43531" + ], + "prepared_review_unit_hash": "984f8bae52e47fe474a4bef743abbcf8453bdf95ec2cf5f379d20f669cfeb9e5", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34753, + "estimated_input_tokens": 8689, + "estimated_eval_tokens": 17634 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a loose set of unrelated Transformers issue reports with superficial keyword overlap (FlashAttention, cache/save/load, multimodal/VL, v5 regressions). I do not see any strong duplicate pairs; all soft edges should be rejected.", + "confidence": 0.44, + "canonical_issue_reason": "issue:43646 is the broadest umbrella-style Transformers v5 regression in the set, so it is the closest single representative, though the cluster is fairly heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is the most concrete and well-scoped core loading bug, with strong discussion/inbound references, making it the best standalone issue to anchor triage if one item must be chosen.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Both involve optional dependencies, but one is a missing utility symbol and the other is a pipelines import failure tied to kernels version/revision requirements; different code paths." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and Mistral-3 image preprocessor dtype mismatch are unrelated bugs in different subsystems." + }, + { + "left": "issue:43316", + "right": "issue:46821", + "accept": false, + "reason": "Config API discrepancy for Gemma3TextConfig is not the same as TimesFM window_size raising AttributeError." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Both mention chunked/compiled execution, but one is inconsistent generation outputs and the other is cache lazy init causing recompilation; not the same bug." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "A broad v5 custom initialization regression is different from GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init validation for patchtsmixer and non-persistent buffer junk-fill are separate v5 issues with different failure modes." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "gguf_file loading breakage and unrecognized-model/model_type errors are both loading problems but not the same underlying code-path." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Gemma-3 eos_token_id inconsistency and sinq quantized save/load breakage concern different model metadata and serialization flows." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "Bare except in a tensor conversion helper and Trainer import failure due to datasets are unrelated." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage and a janus image-generation test failure are different models and different failures." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Weight tying in Mistral3ForConditionalGeneration is a different problem from Ministral-3 FlashAttention compatibility." + }, + { + "left": "issue:44242", + "right": "issue:46894", + "accept": false, + "reason": "Load-balancing loss omission when output_router_logits=False is not the same as the CohereASR double-shift training-loss bug." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and Sam3Video CUDA OOM are not the same issue; one is architectural, the other resource usage." + }, + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "Trust-remote-code cache path collisions and max_cache_length only applying to Quantized Cache are separate cache/config bugs." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "StableLmConfig missing pad_token_idx and a backend mapping error for tensorflow_text are unrelated config vs import/backend issues." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and Qwen2_5_VLProcessor batched chat-template crash have different entrypoints and failure conditions." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "OmDet-Turbo input-size mismatch and Sam3-litetext memory overflow are different symptoms and different model pipelines." + }, + { + "left": "issue:45381", + "right": "issue:47217", + "accept": false, + "reason": "Wrong vision_position_ids for Qwen2.5-VL video inputs is not the same as image-placeholder handling in user text for VL models." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "TensorBoard without TensorFlow causing backend errors and a missing import_utils availability helper are different dependency issues." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "A janus generation test failure and auto_mappings still referencing removed Sam3LiteText configs are separate CI breakages." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy-field loading failure for BEiT semantic segmentation and wrong custom-module restoration after save_pretrained are distinct load-time bugs." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Swallowing KeyboardInterrupt in a tensor conversion helper and a Windows/ROCm torch.distributed import problem are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus test failure and Qwen3.5 FlashAttention generation problems are different models and different code paths." + }, + { + "left": "issue:43352", + "right": "issue:46640", + "accept": false, + "reason": "NemotronHForCausalLM not supporting FlashAttention 2 yet is a specific model capability issue; the other is a generic flash-attention report without evidence of the same root cause." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA merge collapse with extended vocabulary and Gemma 4 mm_token_type_ids defaults are unrelated training/config bugs." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config and Qwen3-MoE sliding_window issues are separate config/runtime 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": 27, + "review_unit_index": 38, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29870", + "issue:35141", + "issue:41950", + "issue:42490", + "issue:43335", + "issue:43388", + "issue:43475", + "issue:43644", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:43827", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44479", + "issue:44534", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44855", + "issue:44869", + "issue:44877", + "issue:44913", + "issue:44964", + "issue:45072", + "issue:45245", + "issue:45406", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45632", + "issue:46082", + "issue:46202", + "issue:46421", + "issue:46617", + "issue:46693", + "issue:46697", + "issue:46832", + "issue:46858", + "issue:46897", + "issue:46932", + "issue:47217" + ], + "soft_pairs": [ + "issue:43650|issue:46202", + "issue:44479|issue:47217", + "issue:44534|issue:44913", + "issue:41950|issue:43475", + "issue:43673|issue:44743", + "issue:35141|issue:42490", + "issue:43388|issue:43688", + "issue:44190|issue:44964", + "issue:46693|issue:47217", + "issue:44855|issue:45588", + "issue:29870|issue:35141", + "issue:45632|issue:46617", + "issue:44869|issue:46832", + "issue:45072|issue:46932", + "issue:45245|issue:46858", + "issue:46082|issue:46697", + "issue:44734|issue:45406", + "issue:43827|issue:43994", + "issue:43644|issue:44913", + "issue:44617|issue:45584", + "issue:25251|issue:27957", + "issue:43335|issue:44855", + "issue:43931|issue:44060", + "issue:44077|issue:44877", + "issue:44242|issue:46897", + "issue:43673|issue:46421", + "issue:45563|issue:45584" + ], + "prepared_review_unit_hash": "3fc661e55383163d92f56ffc4a51bd854b8e5b2c9c15c537f43858f8d9c19f66", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36001, + "estimated_input_tokens": 9001, + "estimated_eval_tokens": 18258 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The proposed soft edges are mostly false positives: the paired issues touch different models, pipelines, cache behaviors, config bugs, or docs, and do not describe the same underlying bug/change. I do not see a clear canonical issue or PR for this set because the items are heterogeneous and none of the candidate pairs are strong duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "Unrelated topics: one is a vague data request, the other is a Gemma pipeline roundtrip tokenizer edge case. No shared code path or concrete duplicate bug." + }, + { + "left": "issue:44479", + "right": "issue:47217", + "accept": false, + "reason": "Different regressions: v5.3.0 video-input model-family issue versus VL placeholder handling in user text. Similar domain, but not the same bug." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Both mention v5/config behavior, but one is about non-persistent buffers being initialized with junk and the other about GPTNeoXConfig rotary_pct not persisting on reload. Different failure modes and code paths." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Video-classification pipeline processor lookup is unrelated to SAM3 vision encoder missing fpn_position_embeddings. Different subsystems and fixes." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Chunked-prefill cache absence in GenerationMixin is not the same as recurrent states resetting in modular_qwen3_5.py. Different mechanisms and models." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "One is token embedding reinitialization after resize; the other is trust_remote_code path normalization with trailing slash. No overlap beyond generic loading behavior." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Both concern training/metrics/losses, but one is gather_for_metrics truncating tuple labels and the other is auxiliary loss normalization in MoE models. Distinct bugs." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading in an image-classification script is unrelated to multimodal model loading failure for Phi-4. Different surfaces and root causes." + }, + { + "left": "issue:46693", + "right": "issue:47217", + "accept": false, + "reason": "Performance regression with FA2 long-context generation is not the same as VL image-placeholder-in-user-text failures. Shared broad area only." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 IndentationError from torch.jit parsing is unrelated to flash_attention.py crashing on s_aux=None. Completely different code paths." + }, + { + "left": "issue:29870", + "right": "issue:35141", + "accept": false, + "reason": "BART generate() decoder_attention_mask bug and embedding reinitialization after resize are different model-generation issues with no common fix." + }, + { + "left": "issue:45632", + "right": "issue:46617", + "accept": false, + "reason": "Both involve trust_remote_code/local cache, but one is a cache-path collision for same leaf names and the other is FileNotFoundError loading from cache. Related area, not same bug." + }, + { + "left": "issue:44869", + "right": "issue:46832", + "accept": false, + "reason": "Whisper trailing replacement-character crash and RT-DETR feature-level mismatch are unrelated decoding vs model-forward issues." + }, + { + "left": "issue:45072", + "right": "issue:46932", + "accept": false, + "reason": "Both mention dtype issues in multimodal models, but one is a CI inference mismatch and the other is quantized Gemma4 vision/audio embedder casting. Not the same code path." + }, + { + "left": "issue:45245", + "right": "issue:46858", + "accept": false, + "reason": "Category-count limit error and device-tensor seq_length arithmetic bug are unrelated. No shared underlying defect." + }, + { + "left": "issue:46082", + "right": "issue:46697", + "accept": false, + "reason": "LlamaConfig head_dim validation and Swin2SR over-padding are different config vs preprocessing bugs." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV cache continuation indexing bug in serve is unrelated to Gemma4Processor missing _tokenizer. Different components and symptoms." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing pipeline() after removals is documentation cleanup, while SigLIP2 AutoModel/pipeline nonsense output is a runtime model/processor issue. Not duplicates." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffers filled with junk is not the same as rotary_pct silently resetting on config reload. Related to serialization only at a high level." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "Sam3Video OOM and Whisper empty-transcription failure are unrelated. One is resource usage, the other is a generation edge case." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "top_k nested-list output bug in a pipeline and XLMRoberta Flash Attention 2 support issue are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "Sparse layer creation bug in SwitchTransformersConfig and a Python 3.13 IndentationError from torch.jit parsing are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch on load and Qwen3-Next tied-weights warning are different loading/config issues with different fixes." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "PatchTSMixer post_init policy and strict config blocking granite_speech are unrelated config-validation concerns." + }, + { + "left": "issue:44242", + "right": "issue:46897", + "accept": false, + "reason": "Load-balancing loss not added when output_router_logits=False is not the same as Florence2 double-shift training-loss bug. Both are loss-related but different defects." + }, + { + "left": "issue:43673", + "right": "issue:46421", + "accept": false, + "reason": "Both concern cache/chunked prefill, but one is missing GenerationMixin cache during chunked_prefill and the other is lazy cache initialization causing recompilation. Different symptoms and likely fixes." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "Paged generate warning about num_return_sequences and Whisper empty-transcription failure are unrelated generation 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": 27, + "review_unit_index": 39, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:25251", + "issue:29870", + "issue:39692", + "issue:41720", + "issue:43065", + "issue:43295", + "issue:43335", + "issue:43526", + "issue:43540", + "issue:43550", + "issue:43643", + "issue:43673", + "issue:43854", + "issue:44112", + "issue:44190", + "issue:44462", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44521", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:44933", + "issue:44938", + "issue:44964", + "issue:45127", + "issue:45245", + "issue:45276", + "issue:45290", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45542", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45907", + "issue:46369", + "issue:46489", + "issue:46655", + "issue:46697", + "issue:46752", + "issue:46858", + "issue:47013" + ], + "soft_pairs": [ + "issue:45381|issue:46369", + "issue:43550|issue:44190", + "issue:41720|issue:45245", + "issue:44933|issue:45362", + "issue:43335|issue:44190", + "issue:44479|issue:44734", + "issue:43065|issue:44610", + "issue:43854|issue:44792", + "issue:45588|issue:47013", + "issue:44617|issue:46369", + "issue:45381|issue:46655", + "issue:39692|issue:44190", + "issue:43295|issue:45276", + "issue:45127|issue:45357", + "issue:44464|issue:45563", + "issue:44779|issue:46489", + "issue:45276|issue:45584", + "issue:43540|issue:44734", + "issue:44112|issue:45600", + "issue:25251|issue:29870", + "issue:45290|issue:46752", + "issue:44462|issue:45907", + "issue:44521|issue:45381", + "issue:43643|issue:44496", + "issue:44938|issue:45542", + "issue:43854|issue:44964", + "issue:43526|issue:46697", + "issue:43673|issue:46858" + ], + "prepared_review_unit_hash": "eb581171d38e55b0ce0a43621c489a910d7bbc41448539d44d285604fca3ff70", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36282, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous issue cluster with no PRs; nearly all pairs are unrelated. The only plausible duplicate-like pair is the DeepSeek tokenizer regression pair (44779/46489).", + "confidence": 0.87, + "canonical_issue_reason": "issue:44779 is the broader, more general report of the DeepSeek tokenizer regression; issue:46489 looks like a more specific instance/gap of that same problem.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44779 is the best anchor for the cluster because it matches the only strong duplicate-like counterpart and describes the broader tokenizer regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "Different models and failures: qwen2.5-vl video vision_position_ids vs Sam3-LiteText memory overflow." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "Unrelated subsystems: Bamba torch.compile/SDPA failure vs local dataset loading in an example script." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Different bug classes: Qwen3 auto device map CUDA assert vs category-count limit error." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "Unrelated issues: missing image_utils import vs Qwen3.5 chat crash." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug is unrelated to dataset-loading failure." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Both are regressions, but one is video-input handling and the other is /v1/responses KV-cache indexing." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Different code paths: dummy Conv2d in Sam3PixelDecoder vs OmDet-Turbo processor/model size mismatch." + }, + { + "left": "issue:43854", + "right": "issue:44792", + "accept": false, + "reason": "Unrelated model-loading problem vs janus test failure." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Different failures in different layers: flash_attention s_aux=None vs xLSTM return shape unpacking." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Both mention memory issues, but they are different models and different symptoms (OOM vs overflow)." + }, + { + "left": "issue:45381", + "right": "issue:46655", + "accept": false, + "reason": "Different multimodal bugs: wrong vision_position_ids vs incorrect padding calculation in SmolVLM video processing." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and quantization issues are unrelated to the local-dataset script bug." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Different APIs and regressions: processor.tokenizer access vs resize_token_embeddings not updating gemma4 embeddings." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA/vocab collapse is unrelated to Qwen3.5 visual-encoder save_pretrained key regression." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Both involve generation, but one is incorrect outputs and the other is a stale warning; not the same bug." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": true, + "reason": "Same underlying DeepSeek tokenizer regression in transformers v5+: both report incorrect tokenizer output, with 46489 describing a specific gap in the same fix." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Gemma4 embedding-resize bug is unrelated to Whisper empty-transcription failure after align_special_tokens." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Qwen3OmniMoe video ValueError and serve KV-cache indexing are different problems." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Separate CI issues affecting different tests and modules." + }, + { + "left": "issue:25251", + "right": "issue:29870", + "accept": false, + "reason": "Pipeline top_k nesting bug and Bart generate() decoder_attention_mask UnboundLocalError are unrelated." + }, + { + "left": "issue:45290", + "right": "issue:46752", + "accept": false, + "reason": "Same function, but different edge cases: tool-call assistant messages vs an empty conversation list." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to list_repo_files not forwarding revision." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal/template issues, but assistant_mask generation and vision_position_ids are distinct code paths." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Different config-loading failures: missing fields under trust_remote_code vs missing model_type in config.json." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import/load failure is unrelated to the tensorboard-without-TF backend error." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different model-loading failures for different models; not the same underlying bug." + }, + { + "left": "issue:43526", + "right": "issue:46697", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Swin2SR over-padding." + }, + { + "left": "issue:43673", + "right": "issue:46858", + "accept": false, + "reason": "Both touch cache/prefill, but one is missing GenerationMixin cache in chunked_prefill and the other is device-tensor slice arithmetic on non-CUDA backends." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Flash-attention sink handling and xLSTM train-mode return unpacking are unrelated failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet: there are no PRs, the items are mostly heterogeneous issues, and 44779/46489 is the only clearly duplicate-like pair. The rejected soft edges are conservative and the accepted DeepSeek tokenizer pair is well supported by the titles and descriptions." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 40, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:27957", + "issue:32090", + "issue:35141", + "issue:36246", + "issue:38617", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:42796", + "issue:43295", + "issue:43381", + "issue:43404", + "issue:43526", + "issue:43531", + "issue:43582", + "issue:43673", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44112", + "issue:44220", + "issue:44246", + "issue:44279", + "issue:44295", + "issue:44336", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44737", + "issue:44964", + "issue:44987", + "issue:45003", + "issue:45042", + "issue:45083", + "issue:45216", + "issue:45335", + "issue:45446", + "issue:45593", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:46082", + "issue:46121", + "issue:46234", + "issue:46424", + "issue:46506", + "issue:46932", + "issue:47217" + ], + "soft_pairs": [ + "issue:44246|issue:45003", + "issue:27957|issue:35141", + "issue:44737|issue:45803", + "issue:44964|issue:45600", + "issue:44112|issue:44964", + "issue:42175|issue:44279", + "issue:44496|issue:47217", + "issue:45632|issue:46234", + "issue:32090|issue:35141", + "issue:42796|issue:44521", + "issue:36246|issue:41628", + "issue:36246|issue:38617", + "issue:43404|issue:45216", + "issue:43531|issue:45600", + "issue:44295|issue:46506", + "issue:43673|issue:46424", + "issue:46082|issue:46121", + "issue:41950|issue:44016", + "issue:43295|issue:45335", + "issue:45083|issue:46506", + "issue:43976|issue:44987", + "issue:44496|issue:45042", + "issue:43582|issue:46932", + "issue:43673|issue:44336", + "issue:43526|issue:43901", + "issue:44683|issue:45446", + "issue:43901|issue:44496", + "issue:44016|issue:44220", + "issue:43381|issue:45593" + ], + "prepared_review_unit_hash": "56147211bfb31b717a2a78a6cd489322d13ff0acf5a88a5519ca9b36a265f49e", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36452, + "estimated_input_tokens": 9113, + "estimated_eval_tokens": 18482 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the soft-similarity pairs mostly share only broad subsystem terms (imports, cache, multimodal, flex_attention, etc.) and do not describe the same concrete bug or change. I would not merge any of the candidate pairs.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Both are transformer internals, but one is a slow import concern and the other is a sys.modules safety bug; different symptoms and code paths." + }, + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "XLMRoberta Flash Attention support and token-embedding reinitialization are unrelated model behaviors." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet CPU tensor creation and a bare-except error swallowing in FuyuBatchFeature are unrelated fixes." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "A Phi-4 multimodal loading failure and stale auto_mappings references to Sam3LiteText are different model/config issues." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flakiness in GraniteSpeech does not match a multimodal model-loading regression." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "Package-install backend dependency omission is not the same as a generic dependency issue report." + }, + { + "left": "issue:44496", + "right": "issue:47217", + "accept": false, + "reason": "Unrecognized model/config loading and VL image-placeholder handling are distinct multimodal loading problems." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "Cache-path collision for trust_remote_code and a missing import_utils helper are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer GPU broadcast TypeError and post_init embedding reinitialization are different bugs." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer BPE marker handling and multimodal assistant_mask generation are separate text-processing issues." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Missing Qwen2.5 VL image processor export and missing AutoImageProcessor import are different missing-symbol problems." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Different missing imports from different modules; no shared concrete failure." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Weight tying in Mistral3 generation and Qwen3.5 save_pretrained checkpoint correctness are unrelated model bugs." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding window behavior and removed Sam3LiteText config references are not the same issue." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "Buffer registration behavior and CPU/CUDA output divergence are different failure modes." + }, + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Chunked prefill cache missing and max_cache_length handling for quantized cache are separate cache-feature bugs." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "LlamaConfig head_dim validation and rope-parameter JSON list handling are unrelated config-validation issues." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification pipeline processor lookup and a notebook syntax error are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Processor/tokenizer multimodal regression and t5gemma token resize not affecting decoder embeddings are different code paths." + }, + { + "left": "issue:45083", + "right": "issue:46506", + "accept": false, + "reason": "A helper-length computation bug in qwen3_omni_moe and CPU/CUDA divergence in PPDocLayoutV3 are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "A Python version compatibility problem and a model-loading failure under transformers>=5.1.0 are not the same defect." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Model loading/config recognition and PIL image processors requiring torchvision are different multimodal issues." + }, + { + "left": "issue:43582", + "right": "issue:46932", + "accept": false, + "reason": "AppleSilicon TypeError in caching_allocator_warmup and Gemma4 dtype-casting fix are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Cache-related GenerationMixin behavior and ANSI codes in loading_report are entirely different concerns." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and docs mentioning return_all_scores are different API/behavior mismatches." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both touch flex_attention, but one is a compiled flex_attention failure on torch>=2.9 and the other is an incorrect version check for AuxRequest import; not the same concrete bug." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Pipeline docs wording and an unrecognized model-loading error are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error and _torch_extract_fbank_features behavior are different issues." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated training/inference 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": 27, + "review_unit_index": 41, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:39692", + "issue:41950", + "issue:42796", + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43577", + "issue:43650", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44220", + "issue:44279", + "issue:44327", + "issue:44351", + "issue:44368", + "issue:44466", + "issue:44779", + "issue:44792", + "issue:44933", + "issue:45020", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45335", + "issue:45357", + "issue:45375", + "issue:45405", + "issue:45584", + "issue:45588", + "issue:45812", + "issue:46082", + "issue:46123", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46515", + "issue:46614", + "issue:46752", + "issue:47000", + "issue:47085", + "issue:47217" + ], + "soft_pairs": [ + "issue:42907|issue:43441", + "issue:43650|issue:44220", + "issue:44016|issue:44062", + "issue:44792|issue:46424", + "issue:43976|issue:44279", + "issue:43441|issue:45588", + "issue:43824|issue:44933", + "issue:44279|issue:46515", + "issue:43479|issue:45375", + "issue:45335|issue:45584", + "issue:43299|issue:45216", + "issue:43901|issue:47217", + "issue:46614|issue:47085", + "issue:44327|issue:46752", + "issue:46123|issue:46369", + "issue:44279|issue:45405", + "issue:43295|issue:44779", + "issue:39692|issue:43901", + "issue:44075|issue:44368", + "issue:43577|issue:45216", + "issue:42796|issue:43650", + "issue:44351|issue:45278", + "issue:41950|issue:45020", + "issue:43976|issue:45405", + "issue:44779|issue:45812", + "issue:43381|issue:47000", + "issue:44466|issue:45276", + "issue:46082|issue:46399", + "issue:43299|issue:45357" + ], + "prepared_review_unit_hash": "a4b832bcc08f7da60d30e1d4e678327cd98e878c4971c8e97cbab6801df3ef2f", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36332, + "estimated_input_tokens": 9083, + "estimated_eval_tokens": 18422 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a collection of unrelated transformer regressions, docs issues, and model-specific bugs; the soft edges are superficial matches on broad themes like \u201cimport error,\u201d \u201ctokenizer,\u201d or \u201cversion regression,\u201d not true duplicates.", + "confidence": 0.18, + "canonical_issue_reason": "Issue 45020 is the broadest umbrella in the set: it describes recent-version breakage affecting models using `remote_code`, which is the closest high-level representative of the cluster\u2019s general regression theme.", + "canonical_pr_reason": null, + "best_issue_reason": "45020 is the best available representative, but it is still too broad and not a true duplicate of the other issues; it should only be used as a loose umbrella, not a merged duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Same model family, but different failures: saving dequantized checkpoints vs FlashAttention loading/runtime crash." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "Unrelated issues: placeholder text vs an audio feature extraction bug." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Docs notebook syntax error is unrelated to the tokenizer `AddedToken` keyword-argument bug." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "Different code paths: Janus image-generation test failure vs cache-length handling in quantized cache." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Generic Python-version compatibility issue vs a dependency problem; too broad to be the same bug." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both touch attention code, but one is a Ministral FlashAttention failure and the other is an integration crash on `s_aux=None`." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import errors, but for different missing symbols/modules and different underlying causes." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "Different dependency incompatibilities: one is a transformers dependency issue, the other is specifically torch 2.4.1." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both involve config handling, but they concern different model families and different missing/defaulted fields." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize propagation is unrelated to Whisper empty-transcription generation after token alignment." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both are Qwen3.5 regressions, but one is model loading and the other is incorrect checkpoint serialization." + }, + { + "left": "issue:43901", + "right": "issue:47217", + "accept": false, + "reason": "Docs mismatch for a pipeline option is unrelated to VL placeholder parsing failures." + }, + { + "left": "issue:46614", + "right": "issue:47085", + "accept": false, + "reason": "Stale phonemizer backend state and tokenizer thread-safety are different subsystems and bugs." + }, + { + "left": "issue:44327", + "right": "issue:46752", + "accept": false, + "reason": "Different edge-case crashes in different utilities; only the generic index-error shape overlaps." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "Batch-finalization bug in generation results vs memory overflow in a specific model call." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Both mention dependencies, but one is an incompatibility issue and the other is an unreleased PEFT version bump." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different failure modes; not the same code-path bug." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Docs issues, but one is a SigLIP2 example/model mismatch and the other is outdated pipeline docs." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer arguments being ignored is unrelated to a tied-embeddings warning during Qwen fine-tuning." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "One is dtype propagation on BLIP2 load, the other is Qwen3.5 checkpoint serialization." + }, + { + "left": "issue:42796", + "right": "issue:43650", + "accept": false, + "reason": "Tokenizer marker handling and a placeholder issue titled \"ADD THE DATA\" are unrelated." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing export (`HybridCache`) is not the same underlying bug as a broad wave of import errors after upgrade." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Different pipeline bug vs broad remote_code regression; too far apart to merge as duplicates." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python compatibility and unreleased PEFT pin are separate dependency problems." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but one is Deepseek-specific incorrect results and the other is Granite token IDs; no clear shared bug." + }, + { + "left": "issue:43381", + "right": "issue:47000", + "accept": false, + "reason": "Gradient-checkpointing eval-mode behavior is unrelated to RTDetr loss infeasible-cost-matrix failure." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Both involve tied/embedding-related model state, but one is serialization behavior and the other is `resize_token_embeddings` propagation." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "Llama config validation and Qwen3.5 kernel registration are different model/config bugs." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5 regressions, but save/load serialization of different components is not 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": 1, + "component_count": 27, + "review_unit_index": 42, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:33290", + "issue:41950", + "issue:42175", + "issue:42994", + "issue:43299", + "issue:43317", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43644", + "issue:43650", + "issue:43653", + "issue:43854", + "issue:43873", + "issue:43931", + "issue:43976", + "issue:44016", + "issue:44242", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44291", + "issue:44336", + "issue:44462", + "issue:44623", + "issue:44683", + "issue:44749", + "issue:45081", + "issue:45103", + "issue:45200", + "issue:45372", + "issue:45397", + "issue:45542", + "issue:45588", + "issue:45593", + "issue:45656", + "issue:45823", + "issue:45834", + "issue:45859", + "issue:45907", + "issue:46082", + "issue:46209", + "issue:46512", + "issue:46851", + "issue:47000", + "issue:47145", + "issue:47217" + ], + "soft_pairs": [ + "issue:43653|issue:45593", + "issue:46082|issue:46512", + "issue:41950|issue:45823", + "issue:44242|issue:45200", + "issue:43644|issue:47145", + "issue:43650|issue:46209", + "issue:43299|issue:43854", + "issue:44462|issue:45081", + "issue:43931|issue:47145", + "issue:44683|issue:45542", + "issue:44016|issue:44291", + "issue:45588|issue:47000", + "issue:43479|issue:43525", + "issue:45081|issue:45907", + "issue:44336|issue:46851", + "issue:42994|issue:43317", + "issue:33290|issue:45656", + "issue:43441|issue:45834", + "issue:44623|issue:44749", + "issue:43873|issue:47145", + "issue:44261|issue:44263", + "issue:45859|issue:46851", + "issue:45103|issue:45372", + "issue:44246|issue:47217", + "issue:44623|issue:45397", + "issue:42175|issue:43976" + ], + "prepared_review_unit_hash": "ba2209d00ad7c9f3aa4c03e3ab782760622374f269764f193cac89b76ab89f03", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34840, + "estimated_input_tokens": 8710, + "estimated_eval_tokens": 17676 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the items cover unrelated tokenizer, multimodal config, quantization/offload, deepspeed, model-loading, and docs/notebook bugs. None of the soft pairs look like the same underlying defect, so no duplicate merge is justified.", + "confidence": 0.95, + "canonical_issue_reason": "No canonical issue stands out because the cluster does not share one underlying bug or change; the issues span many unrelated subsystems and model families.", + "canonical_pr_reason": "No PRs are present, so there is no canonical PR.", + "best_issue_reason": "No single issue is a good global representative: any pick would be arbitrary given the lack of a common failure mode across the set.", + "best_pr_reason": "No PRs are present, so there is no best PR candidate.", + "soft_edge_verdicts": [ + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "BigBirdTokenizer special-token decoding and D-FINE auxiliary-loss behavior are unrelated subsystems with different symptoms." + }, + { + "left": "issue:46082", + "right": "issue:46512", + "accept": false, + "reason": "LlamaConfig head_dim validation and from_config(dtype) stale config.dtype are different config-handling bugs." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Pipeline processor lookup and Gemma4 device-map forward mismatch are not the same code-path problem." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Missing load-balancing loss and Gemma4 mm_token_type_ids defaults are separate model-training issues." + }, + { + "left": "issue:43644", + "right": "issue:47145", + "accept": false, + "reason": "Both involve loading/state handling, but one is non-persistent buffers and the other is FP8 weight_scale application; different root causes." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Placeholder title aside, this is unrelated to FP8 inference on B200." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different model families and different loading failures; not one shared bug." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashes are separate tokenizer-loading paths." + }, + { + "left": "issue:43931", + "right": "issue:47145", + "accept": false, + "reason": "Weight-shape mismatch for Qwen3-VL and FP8 frozen-checkpoint scaling are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Compiled flex_attention on torch 2.9 and tensorboard-without-TensorFlow backend errors have no shared underlying defect." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error and init_empty_weights argument mismatch are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:47000", + "accept": false, + "reason": "Flash-attention AttributeError on s_aux=None and RTDetrLoss infeasible cost matrix are different model/runtime bugs." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Both are config issues, but they affect different models and different fields; not the same bug." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Mistral tokenizer regex patch crash and list_repo_files revision forwarding are unrelated." + }, + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI escape emission in loading_report and Qwen3.5 variable-length position_ids are unrelated." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Quantized model saving failure and dequantized-model offload loading failure are distinct issues." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Deepspeed Adafactor OOM and deepspeed optimizer.step being called twice are different bugs despite the shared framework." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "Ministral FlashAttention failure and Kosmos2.5 long-OCR index error are unrelated." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "processor.save_pretrained missing files and import slowdown after upgrade are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:47145", + "accept": false, + "reason": "General quantization offloading behavior and FP8 weight_scale omission are different problems." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Both are GLM-related, but one is a missing rms_norm_eps default and the other is a torch.split indexing issue." + }, + { + "left": "issue:45859", + "right": "issue:46851", + "accept": false, + "reason": "CPU offload incompatibility in a Qwen3.5 rotary embedding and variable-length position_ids are not the same failure mode." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Auto-docstring annotation handling and Gemma 4 processor import failure are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:47217", + "accept": false, + "reason": "Import-time slowness and VL image-placeholder handling are unrelated." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Missing processor save files and gemma-4 zero3 from_pretrained are distinct 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": 27, + "review_unit_index": 43, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:41628", + "issue:42796", + "issue:43441", + "issue:43525", + "issue:43526", + "issue:43575", + "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:44297", + "issue:44336", + "issue:44373", + "issue:44496", + "issue:44589", + "issue:44671", + "issue:44779", + "issue:44871", + "issue:44987", + "issue:45003", + "issue:45083", + "issue:45325", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45600", + "issue:45712", + "issue:45823", + "issue:46291", + "issue:46399", + "issue:46655", + "issue:46726", + "issue:46851", + "issue:47000", + "issue:47145", + "issue:47176", + "issue:47177", + "issue:47217" + ], + "soft_pairs": [ + "issue:47176|issue:47177", + "issue:45600|issue:45712", + "issue:44671|issue:44779", + "issue:43525|issue:43881", + "issue:43901|issue:44246", + "issue:43673|issue:43761", + "issue:44188|issue:44336", + "issue:43575|issue:47145", + "issue:44016|issue:45823", + "issue:44186|issue:47000", + "issue:43824|issue:43976", + "issue:44987|issue:45542", + "issue:43525|issue:43643", + "issue:44336|issue:44373", + "issue:43723|issue:44162", + "issue:42796|issue:43441", + "issue:44589|issue:45362", + "issue:45325|issue:46851", + "issue:46291|issue:46399", + "issue:44297|issue:45397", + "issue:44246|issue:44496", + "issue:45083|issue:46851", + "issue:44871|issue:47145", + "issue:41628|issue:43976", + "issue:46655|issue:46726", + "issue:43784|issue:45542", + "issue:45003|issue:47217", + "issue:46399|issue:46851", + "issue:43526|issue:45003" + ], + "prepared_review_unit_hash": "3abc04b3ac4d9544c83ffc68af38a1f58e2576a6d5c568090891c8a3892cb41c", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36341, + "estimated_input_tokens": 9086, + "estimated_eval_tokens": 18428 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose grab-bag of mostly unrelated Transformers issues: model/tokenizer loading regressions, config mismatches, processor bugs, kernel/runtime failures, docs typos, and a couple security reports. There is no strong duplicate center.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue is a good canonical representative because the items are too heterogeneous and do not describe one underlying bug.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a true global best fit; if one had to be picked, issue:43723 is the broadest user-facing loading regression, but it still does not unify the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47176", + "right": "issue:47177", + "accept": false, + "reason": "Different security bugs: command-substitution allowlist bypass vs tar/zip slip arbitrary file write." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Both are repo hygiene problems, but one is stale auto_mappings config refs and the other is dummy classes leaking into __dir__/check_repo." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "Different models and failures: CamemBERT MLM predictions vs Deepseek tokenizer output corruption." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Different loading failures with different root causes; only the symptom is a generic model-load crash." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Docs mismatch for return_all_scores is unrelated to intermittent import slowness." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Both are v5 regressions, but one is chunked_prefill cache handling and the other is CLIPVision hidden_states propagation." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel branching under torch.compile is unrelated to ANSI escapes in loading_report." + }, + { + "left": "issue:43575", + "right": "issue:47145", + "accept": false, + "reason": "Tensor-parallel OOM while loading Qwen2-57B is unrelated to FP8 frozen-checkpoint weight_scale handling." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "Notebook syntax error and Gemma4 forward device mismatch are different problems." + }, + { + "left": "issue:44186", + "right": "issue:47000", + "accept": false, + "reason": "LayoutLMv2 tokenizer padding/truncation crash is unrelated to RTDetrLoss cost-matrix infeasibility." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Qwen2_5_VL import error and Python version compatibility failure are separate issues." + }, + { + "left": "issue:44987", + "right": "issue:45542", + "accept": false, + "reason": "Model loading failure on v5.1.0 and missing TensorFlow backend when only tensorboard is installed are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Both mention missing fields, but one is a specific Llama4Config attribute bug and the other is trust_remote_code AutoConfig field loss." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI code leakage in loading_report is unrelated to a wrong docstring." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "Tokenizer loading regression and ESM2 breakage are different model-specific failures." + }, + { + "left": "issue:42796", + "right": "issue:43441", + "accept": false, + "reason": "Ministral tokenizer BPE-marker handling and FlashAttention failure are distinct bugs." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage lookup failure is unrelated to Qwen3.5 chat loading/crash behavior." + }, + { + "left": "issue:45325", + "right": "issue:46851", + "accept": false, + "reason": "Both concern position_ids, but they affect different model families and different logic paths." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "kernels import/version requirement failure is unrelated to a missing decorator on apply_rotary_pos_emb registration." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer config class mismatch is not the same bug as gemma-4 zero3 from_pretrained failure." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import latency and unrecognized model config are unrelated." + }, + { + "left": "issue:45083", + "right": "issue:46851", + "accept": false, + "reason": "Different Qwen-family position handling bugs in different code paths." + }, + { + "left": "issue:44871", + "right": "issue:47145", + "accept": false, + "reason": "Gemma eos_token_id inconsistency is unrelated to FP8 weight_scale loading." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "AutoImageProcessor import failure and Python 5.1.0 compatibility are not the same issue." + }, + { + "left": "issue:46655", + "right": "issue:46726", + "accept": false, + "reason": "Video padding miscalculation and missing-vision-backend image processor error are different processor problems." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "sentence-transformers import NameError and undefined tf backend error have different causes." + }, + { + "left": "issue:45003", + "right": "issue:47217", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to VL placeholder handling in user text." + }, + { + "left": "issue:46399", + "right": "issue:46851", + "accept": false, + "reason": "Kernel registration crash and variable-length position_ids bug are different Qwen3.5 issues." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to modeling_utils sys.modules access." + } + ] + }, + "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": 27, + "review_unit_index": 44, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:36246", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43257", + "issue:43299", + "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:44749", + "issue:44779", + "issue:44871", + "issue:45003", + "issue:45072", + "issue:45103", + "issue:45137", + "issue:45292", + "issue:45362", + "issue:45375", + "issue:45397", + "issue:45561", + "issue:45563", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:46132", + "issue:46399", + "issue:46506", + "issue:46614", + "issue:46655", + "issue:46665", + "issue:46726", + "issue:46752", + "issue:46821", + "issue:46829", + "issue:46851", + "issue:47217" + ], + "soft_pairs": [ + "issue:43540|issue:46752", + "issue:45103|issue:46399", + "issue:45072|issue:46829", + "issue:45561|issue:45563", + "issue:44295|issue:45397", + "issue:43531|issue:45362", + "issue:44246|issue:44871", + "issue:44336|issue:44485", + "issue:45588|issue:46851", + "issue:46614|issue:46665", + "issue:42491|issue:43299", + "issue:45003|issue:46821", + "issue:44242|issue:45593", + "issue:43550|issue:46506", + "issue:42886|issue:46132", + "issue:36246|issue:42757", + "issue:44779|issue:45292", + "issue:36246|issue:44351", + "issue:46726|issue:47217", + "issue:43901|issue:45003", + "issue:43257|issue:45137", + "issue:41950|issue:44220", + "issue:43540|issue:46655", + "issue:44451|issue:46752", + "issue:45375|issue:45759", + "issue:44479|issue:46655", + "issue:44295|issue:44749" + ], + "prepared_review_unit_hash": "5c8ca2f0553b227f3e233ce5596fc0d113cf13d74db878b5b9a048b97bb8a6f3", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35768, + "estimated_input_tokens": 8942, + "estimated_eval_tokens": 18140 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: several Qwen/VL/video/loading regressions are mixed with unrelated import, docs, performance, and training issues. I did not find any pair that clearly looks like the same underlying bug, so all soft edges should be rejected. ", + "confidence": 0.67, + "canonical_issue_reason": "issue:44479 is the broadest and most explicit regression report in the set, spanning multiple Qwen VL/MoE variants with a clear version boundary (v5.3.0) and a concrete failure mode. If one issue must anchor this mixed cluster, it is the best representative.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the most reusable cluster representative because it describes a concrete, multi-model regression rather than a narrow one-off symptom. The rest of the items are either unrelated or too specific to serve as a stable umbrella.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43540", + "right": "issue:46752", + "accept": false, + "reason": "Both are runtime failures, but one is Qwen3OmniMoe video-input handling and the other is empty-chat-template handling; different code paths and causes." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "One is an auto-docstring annotation crash, the other is a Qwen3.5 kernel-registration error; unrelated subsystems." + }, + { + "left": "issue:45072", + "right": "issue:46829", + "accept": false, + "reason": "bfloat16 dtype mismatch in inference vs incorrect hidden_states assignment in a video-classification output; not the same bug." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "pytest-xdist file race and stale paged-generate warning are unrelated behavior issues." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer access vs gemma-4 zero3 from_pretrained failure; different failure mechanisms." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Qwen3-MoE sliding_window handling and Qwen3.5 chat crash are distinct model-path problems." + }, + { + "left": "issue:44242", + "right": "issue:44871", + "accept": false, + "reason": "Missing load-balancing loss and inconsistent eos_token_id config affect different parts of different models." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "Terminal-color output bug in loading_report vs GLM-5 RoPE implementation discussion; no shared underlying defect." + }, + { + "left": "issue:45588", + "right": "issue:46851", + "accept": false, + "reason": "Flash-attention sinkless-model crash and variable-length position_ids issue are separate runtime bugs." + }, + { + "left": "issue:46614", + "right": "issue:46665", + "accept": false, + "reason": "Stale phonemizer backend state vs from_pretrained behavior inside Python classes; unrelated." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both involve Qwen3 MoE, but one is a conversion/deepspeed issue and the other is a loading regression in v5; not the same concrete bug." + }, + { + "left": "issue:45003", + "right": "issue:46821", + "accept": false, + "reason": "Unsafe sys.modules access and TimesFM window_size AttributeError are unrelated implementation bugs." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Both mention losses, but they are different models and different missing-loss conditions." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "torch.compile/SDPA failure for Bamba and CPU/CUDA output divergence for PPDocLayoutV3 are distinct issues." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and AttentionInterface.register behavior change are unrelated." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different symbols and different root causes." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Tokenizer correctness regression and resize_token_embeddings not updating output embeddings are different bugs." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports in different modules; same symptom class but not the same underlying issue." + }, + { + "left": "issue:46726", + "right": "issue:47217", + "accept": false, + "reason": "Missing vision backend handling vs image placeholder-in-text failure are separate VL issues." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Documentation mismatch and unsafe sys.modules access are unrelated." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both involve DeepSpeed, but one is model weight conversion and the other is a ZeRO3 deque error; not mergeable as one fix." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification pipeline image-processor lookup and fbank feature extraction are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46655", + "accept": false, + "reason": "Both touch video processing, but they are different models and different defects (input parsing vs padding calculation)." + }, + { + "left": "issue:44451", + "right": "issue:46752", + "accept": false, + "reason": "ScandiBERT loading failure and empty-list chat template crash are unrelated." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Missing config field under @strict and unwrapping text_config for composite multimodal configs are different loading/config bugs." + }, + { + "left": "issue:44479", + "right": "issue:46655", + "accept": false, + "reason": "Both are video-related, but one is a broad Qwen VL regression and the other is a SmolVLM padding bug; different code paths." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Position_ids buffer error and a performance regression after upgrade are not the same defect." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 45, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:36246", + "issue:42199", + "issue:43065", + "issue:43295", + "issue:43352", + "issue:43408", + "issue:43475", + "issue:43526", + "issue:43540", + "issue:43596", + "issue:43653", + "issue:44038", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44623", + "issue:44857", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45250", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45459", + "issue:45561", + "issue:45600", + "issue:45656", + "issue:45859", + "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:45292|issue:47000", + "issue:45859|issue:46399", + "issue:45137|issue:45600", + "issue:43295|issue:45292", + "issue:45003|issue:45561", + "issue:36246|issue:43596", + "issue:43475|issue:46369", + "issue:44466|issue:45335", + "issue:44623|issue:46726", + "issue:44038|issue:46506", + "issue:43526|issue:46897", + "issue:44877|issue:45375", + "issue:44242|issue:44857", + "issue:44987|issue:45459", + "issue:29870|issue:30333", + "issue:43352|issue:45250", + "issue:43526|issue:46894", + "issue:44479|issue:46170", + "issue:42199|issue:46851", + "issue:45397|issue:45656", + "issue:43653|issue:44162", + "issue:44368|issue:44977", + "issue:44060|issue:44077", + "issue:43065|issue:43408", + "issue:44877|issue:46170", + "issue:43475|issue:43540" + ], + "prepared_review_unit_hash": "91f959f1d121b7e903173fea5162cc4a6ca8e560bfe26af2de84333ab1bcae04", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35894, + "estimated_input_tokens": 8974, + "estimated_eval_tokens": 18204 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most items are unrelated model-specific regressions, config issues, or infrastructure bugs. None of the soft pairs looks like the same underlying fix, so there is no strong duplicate hub.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue cleanly represents the cluster; the titles span unrelated bugs across different models, runtimes, and subsystems.", + "canonical_pr_reason": null, + "best_issue_reason": "If a representative issue must be chosen, 44479 is the broadest user-facing regression in the set because it impacts several Qwen video variants; still, this is not a true duplicate anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "Different models and failures: ESM2 is broken vs Gemma 4 needs mm_token_type_ids defaults." + }, + { + "left": "issue:45292", + "right": "issue:47000", + "accept": false, + "reason": "Unrelated code paths: token embedding resize bug vs RTDetr loss infeasible cost matrix." + }, + { + "left": "issue:45859", + "right": "issue:46399", + "accept": false, + "reason": "Both touch Qwen3.5, but one is CPU offload rotary embedding and the other is a missing kernel decorator." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque crash vs stale auto_mappings config references; distinct bugs." + }, + { + "left": "issue:43295", + "right": "issue:45292", + "accept": false, + "reason": "Processor/tokenizer regression vs embedding resize propagation; different components and symptoms." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "sys.modules access bug vs pytest-xdist file race; unrelated issues." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Missing Qwen2_5_VL import vs ZeRO3/BertModel IndexError; no shared root cause." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "SAM3 Video missing attribute vs SAM3 LiteText memory overflow; different failures." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both involve tied/embedded weights, but one is serialization and the other is resize propagation in T5Gemma." + }, + { + "left": "issue:44623", + "right": "issue:46726", + "accept": false, + "reason": "save_pretrained missing files vs backend-missing AutoProcessor warning; distinct processor problems." + }, + { + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3-VL-Moe loading regression vs CPU/CUDA result mismatch in PPDocLayoutV3." + }, + { + "left": "issue:43526", + "right": "issue:46897", + "accept": false, + "reason": "Different label-handling bugs in different models; not the same underlying change." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both are strict-config complaints, but one blocks granite_speech loading while the other drops a Qwen field." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Router/load-balancing loss gating vs LwDetr float16 AMP crash; unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Physical-intelligence loading failure and protobuf tokenizer error masking are separate issues." + }, + { + "left": "issue:29870", + "right": "issue:30333", + "accept": false, + "reason": "Bart generate UnboundLocalError vs MLflow job reporting failure; unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Generic flash-attention topic vs a specific Nemotron FA2 support error; insufficient evidence of same bug." + }, + { + "left": "issue:43526", + "right": "issue:46894", + "accept": false, + "reason": "Reduce-labels truncation vs ASR training-loss double-shift; similar pattern only at a very high level." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Video input regression in Qwen models vs Windows ROCm/torch.distributed import issue; distinct code paths." + }, + { + "left": "issue:42199", + "right": "issue:46851", + "accept": false, + "reason": "DETR cardinality error vs Qwen3.5 variable-length position_ids bug; unrelated." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma-4 Zero3 from_pretrained problem vs optimizer stepping twice under DeepSpeed; different runtime bugs." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 breakage; unrelated." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "Tie-word-embeddings warning during LoRA fine-tuning vs flash-attention generation failure; different issues." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Qwen3-Next tied-weights warning vs patchtsmixer post_init policy change; unrelated." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 pixel decoder dummy Conv2d vs sam3_video/sam3_tracker config mismatch; not the same bug." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict-config loading issue vs Windows ROCm torch.distributed dependency problem; unrelated." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "SAM3 Video missing attribute vs Qwen3OmniMoe video input ValueError; different models and 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": 27, + "review_unit_index": 46, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:42886", + "issue:43065", + "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:44964", + "issue:44995", + "issue:45103", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45375", + "issue:45412", + "issue:45478", + "issue:45563", + "issue:45588", + "issue:45593", + "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:46832", + "issue:46858", + "issue:47020", + "issue:47068", + "issue:47145" + ], + "soft_pairs": [ + "issue:45412|issue:46123", + "issue:42886|issue:45229", + "issue:46170|issue:46832", + "issue:43723|issue:46710", + "issue:43931|issue:44368", + "issue:44188|issue:46202", + "issue:44995|issue:45200", + "issue:44368|issue:45563", + "issue:46132|issue:47145", + "issue:44483|issue:46202", + "issue:44188|issue:46421", + "issue:43906|issue:46202", + "issue:45593|issue:47068", + "issue:44995|issue:45588", + "issue:45987|issue:46082", + "issue:45685|issue:46431", + "issue:44361|issue:44749", + "issue:44297|issue:44623", + "issue:45103|issue:46858", + "issue:43065|issue:46123", + "issue:43701|issue:44749", + "issue:44749|issue:45656", + "issue:43994|issue:46726", + "issue:43646|issue:44162", + "issue:44623|issue:45656", + "issue:44964|issue:45137", + "issue:44871|issue:46132", + "issue:46752|issue:47020", + "issue:45375|issue:45478" + ], + "prepared_review_unit_hash": "44150872da82432727b80c397b87a6d5eff14f09e363b1b611675f60148d68d0", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36046, + "estimated_input_tokens": 9012, + "estimated_eval_tokens": 18280 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the issues span tokenizer loading, model/config regressions, training/optimizer behavior, cache semantics, and unrelated subsystem bugs. Only one soft pair looks like a true duplicate/near-duplicate: the StaticCache.get_seq_length contract bug and its _prefill slice-arithmetic failure are the same underlying issue.", + "confidence": 0.73, + "canonical_issue_reason": "issue:45987 is the clearest root-cause report: it names the API contract violation in StaticCache.get_seq_length, which directly explains the downstream failure described in issue:46858.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45987 is the best anchor for the only convincing duplicate pair because it states the underlying bug more generally, while issue:46858 is a specific manifestation.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "Different subsystems and failures: RT-DETR memory retention vs MaskGenerationPipeline partial-batch result dropping." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Tokenizer offline cache loading is unrelated to Gemma4 multi-GPU CUDA OOM." + }, + { + "left": "issue:46170", + "right": "issue:46832", + "accept": false, + "reason": "Windows/ROCm import dependency issue is unrelated to an RTDetrModel forward shape mismatch." + }, + { + "left": "issue:43723", + "right": "issue:46710", + "accept": false, + "reason": "Both are version-related regressions, but one is tokenizer loading and the other is accuracy regression; not the same bug." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch is unrelated to a warning about tie_word_embeddings in Qwen3.5 LoRA fine-tuning." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "Different concrete problems: torch.compile attention-kernel divergence vs Gemma pipeline roundtrip failure." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "GlmMoeDsa stale cache crash is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A configuration warning during fine-tuning is unrelated to a stale generate warning for num_return_sequences." + }, + { + "left": "issue:46132", + "right": "issue:47145", + "accept": false, + "reason": "AttentionInterface.register behavior change is unrelated to FP8 checkpoint loading missing weight_scale application." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "Chat-completions API request rejection is unrelated to Gemma pipeline token roundtrip failure." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "Both mention compile/cache behavior, but one is attention-kernel branching and the other is cache lazy-init recompilation; not the same defect." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "An isolated reproduction of another issue is too generic to match the specific Gemma '\u2581' roundtrip failure." + }, + { + "left": "issue:45593", + "right": "issue:47068", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior and a deepspeed host-synchronization concern are different training-path bugs." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Second-forward-pass cache corruption in GlmMoeDsa is unrelated to flash_attention crashing on s_aux=None." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "Cache sequence-length type violation is unrelated to LlamaConfig rejecting an explicit head_dim." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "MPS histogram support failure is unrelated to missing chat-template errors in processors." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Tokenizer AttributeError on tasks is unrelated to a training throughput regression after upgrading Transformers." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both are save_pretrained-related, but one is tokenizer class metadata mismatch and the other is missing processor files; not the same concrete bug." + }, + { + "left": "issue:45103", + "right": "issue:46858", + "accept": false, + "reason": "auto_docstring annotation handling is unrelated to cache.get_seq_length returning a device tensor in prefill." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to the last-batch handling bug in MaskGenerationPipeline." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to a performance regression after version upgrade." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Throughput slowdown and double optimizer.step under deepspeed are different training issues." + }, + { + "left": "issue:43994", + "right": "issue:46726", + "accept": false, + "reason": "SigLIP2 incorrect outputs and missing vision backend handling are different vision-stack problems." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Custom model initialization breakage in v5 and ESM2 being broken are not the same code-path issue." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "processor.save_pretrained file emission and deepspeed optimizer-step duplication are unrelated." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading failure is unrelated to DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma-3 eos_token_id mismatch is unrelated to AttentionInterface.register behavior changes." + }, + { + "left": "issue:46752", + "right": "issue:47020", + "accept": false, + "reason": "Empty-list chat-template IndexError and XLMRobertaTokenizer/Unigram constructor mismatch are different tokenizer bugs." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 MoE, but one is a missing config field dropped by strict validation and the other is a broader from_pretrained error; insufficiently specific to merge." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": true, + "reason": "Same underlying cache API bug: get_seq_length returns a tensor instead of an int, and 46858 is the concrete _prefill failure this causes on non-CUDA backends." + } + ] + }, + "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": 27, + "review_unit_index": 47, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:36010", + "issue:39692", + "issue:41628", + "issue:42371", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43673", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43873", + "issue:44038", + "issue:44188", + "issue:44222", + "issue:44351", + "issue:44373", + "issue:44623", + "issue:44661", + "issue:44857", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:45092", + "issue:45137", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45397", + "issue:45561", + "issue:45593", + "issue:45753", + "issue:46326", + "issue:46399", + "issue:46463", + "issue:46726", + "issue:46821", + "issue:46851", + "issue:46877", + "issue:47000", + "issue:47059", + "issue:47217" + ], + "soft_pairs": [ + "issue:43475|issue:46821", + "issue:45561|issue:47217", + "issue:42994|issue:43482", + "issue:43475|issue:43643", + "issue:44188|issue:46851", + "issue:45230|issue:45397", + "issue:44373|issue:46851", + "issue:43232|issue:43673", + "issue:45310|issue:45375", + "issue:28282|issue:30990", + "issue:44038|issue:44623", + "issue:39692|issue:46726", + "issue:41628|issue:44933", + "issue:46877|issue:47000", + "issue:46463|issue:46877", + "issue:42994|issue:43525", + "issue:44869|issue:45356", + "issue:43723|issue:46326", + "issue:42371|issue:45092", + "issue:43873|issue:44945", + "issue:43784|issue:46399", + "issue:45137|issue:47000", + "issue:41628|issue:44351", + "issue:36010|issue:41628", + "issue:42994|issue:43824", + "issue:44038|issue:45310", + "issue:44945|issue:45753", + "issue:45561|issue:47059", + "issue:44857|issue:46877", + "issue:44222|issue:45310", + "issue:44661|issue:47059", + "issue:45593|issue:46877" + ], + "prepared_review_unit_hash": "5d5c2d822108bac561f4b58ad5a7a5a04e834ca91f3bf923d864288105b48853", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36847, + "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 mostly heterogeneous: many unrelated import, generation, quantization, vision, and testing regressions got linked only by broad Transformers v5 similarity. I found one plausible duplicate-style pair in the Qwen3.5 MoE loading/config path, but the rest should stay separate.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45310 is the broadest user-facing report in the only clearly related subthread: a Qwen3.5 MoE from_pretrained regression on transformers>=5.4.0. It is a better umbrella than the more specific config-field report.", + "canonical_pr_reason": null, + "best_issue_reason": "45310 is the best global issue anchor because it names the affected model family and the concrete loading regression, and it can subsume the more specific missing-field symptom in 45375.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43475", + "right": "issue:46821", + "accept": false, + "reason": "Both are AttributeError reports, but they hit different models, different attributes, and different code paths." + }, + { + "left": "issue:45561", + "right": "issue:47217", + "accept": false, + "reason": "One is a pytest-xdist/test utility race; the other is a VL prompt/image-placeholder inference bug." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Both involve quantization-related workflows, but one is saving and the other is loading a different model format; not the same bug." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "SAM3 encoder attribute failure and AutoConfig missing fields are unrelated symptoms with different components." + }, + { + "left": "issue:44188", + "right": "issue:46851", + "accept": false, + "reason": "Different subsystems and failure modes: torch.compile attention-kernel divergence vs Qwen3.5 variable-length position_ids handling." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Too generic to merge: one is a placeholder bug report, the other is a specific gemma-4 ZeRO3 from_pretrained issue." + }, + { + "left": "issue:44373", + "right": "issue:46851", + "accept": false, + "reason": "A wrong docstring for position_ids is not the same as a runtime failure in variable-length Qwen3.5 inputs." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both mention generation internals, but they describe different failures in different generation paths." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": true, + "reason": "Same underlying Qwen3.5 MoE loading/config regression: 45310 reports the v5.4.0 from_pretrained failure, while 45375 identifies the missing deepstack_visual_indexes field dropped by strict config handling." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "PyTorch-missing import error and Sentence Transformers hanging on load are different problems." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe bug and processor.save_pretrained missing files are separate issues." + }, + { + "left": "issue:39692", + "right": "issue:46726", + "accept": false, + "reason": "Both touch vision/processor docs, but one is a broken example and the other is an AutoProcessor backend-missing error." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they concern different symbols and likely different export paths." + }, + { + "left": "issue:46877", + "right": "issue:47000", + "accept": false, + "reason": "Import-time docstring lint output and RTDetrLoss infeasible cost matrix are unrelated." + }, + { + "left": "issue:46463", + "right": "issue:46877", + "accept": false, + "reason": "Incorrect T5 validation messaging is not the same bug as doc-lint output at import time." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized save failure and Llama4Config pad_token_id absence are different model/config bugs." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenization/decoding, but they affect different models and different decoding failures." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer loading in v5 and mistral_common continue_final_message handling are separate code paths." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 settings guidance and InternVL2 meta-initialization incompatibility are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading and pipeline-parallel LLM output divergence are different runtime problems." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "sentence-transformers import NameError and Qwen3.5 kernel registration failure are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:47000", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque underflow and RTDetrLoss infeasible cost matrix are different failures." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are top-level import errors, but for different symbols and likely different compatibility fixes." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "GenerationMixin import path breakage and AutoImageProcessor import failure are separate API regressions." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized save failure and missing Qwen2.5-VL class import are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Different Qwen model variants and different failure points; too broad to treat as one bug." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline-parallel output errors and Qwen3.5 image-specific looping are not the same code-path issue." + }, + { + "left": "issue:45561", + "right": "issue:47059", + "accept": false, + "reason": "A test race on captured_info.txt is unrelated to a huggingface_hub CLI compatibility break." + }, + { + "left": "issue:44857", + "right": "issue:46877", + "accept": false, + "reason": "Float16 AMP CUDA crash in LwDetrLoss is unrelated to import-time docstring diagnostics." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 save_pretrained for MoE and Qwen3.5 from_pretrained loading failure are different operations and likely different fixes." + }, + { + "left": "issue:44661", + "right": "issue:47059", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES model-generation failure and CLI completion regression are unrelated." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "D-FINE auxiliary-loss logic and doc-lint stdout noise are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats the cluster as mostly heterogeneous and only elevates the 45310/45375 Qwen3.5 MoE pair as a plausible duplicate-style match. The other soft-edge rejections are cautious and supported by the issue titles/descriptions rather than overstated evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 48, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:39692", + "issue:42175", + "issue:42673", + "issue:42831", + "issue:42898", + "issue:43232", + "issue:43479", + "issue:43704", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43976", + "issue:44016", + "issue:44186", + "issue:44222", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44743", + "issue:44779", + "issue:44792", + "issue:44869", + "issue:44933", + "issue:45081", + "issue:45103", + "issue:45310", + "issue:45405", + "issue:45406", + "issue:45561", + "issue:45600", + "issue:45712", + "issue:45812", + "issue:46077", + "issue:46123", + "issue:46202", + "issue:46231", + "issue:46693", + "issue:46701", + "issue:46829", + "issue:46877", + "issue:46909", + "issue:47059", + "issue:47068", + "issue:47085" + ], + "soft_pairs": [ + "issue:45406|issue:46202", + "issue:45081|issue:45103", + "issue:46202|issue:46701", + "issue:43232|issue:44792", + "issue:42673|issue:43704", + "issue:42831|issue:44222", + "issue:42175|issue:44933", + "issue:45405|issue:46877", + "issue:44743|issue:47068", + "issue:44496|issue:45561", + "issue:30333|issue:33290", + "issue:43976|issue:44933", + "issue:30990|issue:35141", + "issue:42898|issue:44568", + "issue:42175|issue:45712", + "issue:43479|issue:46829", + "issue:44568|issue:47085", + "issue:44623|issue:45310", + "issue:42673|issue:46693", + "issue:44186|issue:47085", + "issue:46077|issue:46123", + "issue:45600|issue:47059", + "issue:42898|issue:45812", + "issue:43723|issue:44568", + "issue:43479|issue:44869", + "issue:44016|issue:46231", + "issue:39692|issue:46909", + "issue:43784|issue:43824", + "issue:42898|issue:44779" + ], + "prepared_review_unit_hash": "23dcb6de393894fe5ae26128f6873d120ae916b9062cf47874102e89307772ad", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36605, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Only one soft edge is a clear duplicate: the Qwen3ForCausalLM VRAM leak report appears twice with the same title. The other candidate pairs are mostly broad-theme matches (tokenizers, Gemma, FP8, docs, generation) but describe different concrete failures and should stay separate.", + "confidence": 0.96, + "canonical_issue_reason": "Use issue:42673 as canonical: it is the earlier, fuller report of the exact same Qwen3ForCausalLM multi-dataloader VRAM leak, and issue:43704 is a near-verbatim duplicate with less context.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the best issue fit because it most clearly states the bug, has the richer discussion trail, and cleanly represents the duplicated Qwen3ForCausalLM VRAM leak problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Both involve Gemma, but one is a processor init bug and the other is a pipeline roundtrip/tokenization failure; different code paths and symptoms." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Both are AttributeErrors, but one is a Mistral tokenizer regex patch failure and the other is auto-docstring handling of future annotations; unrelated bugs." + }, + { + "left": "issue:46202", + "right": "issue:46701", + "accept": false, + "reason": "Gemma-related, but one is a tokenizer/pipeline roundtrip issue and the other is a DiffusionGemmaGenerationOutput pipeline failure; not the same defect." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs handling after sync_gpus is unrelated to the janus image generation test failure." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Same title and same underlying symptom: Qwen3ForCausalLM leaks VRAM when used in multiple dataloader threads." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "Both mention FP8, but one is an accuracy regression and the other is a save_pretrained/moe serialization issue." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "TensorFlow backend packaging and a missing image_utils import are unrelated." + }, + { + "left": "issue:45405", + "right": "issue:46877", + "accept": false, + "reason": "A dependency version bump issue and an auto_docstring stdout logging issue are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:47068", + "accept": false, + "reason": "A Qwen3.5 recurrent-state caching bug and a deepspeed loss synchronization concern are different code paths." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Model loading/config recognition and a pytest-xdist file race are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow reporting on failed training jobs is unrelated to a Sentence Transformers loading hang." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version support and a missing image_utils import are separate problems." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "A loading hang and an embedding reinitialization bug are different failures." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one concerns clean_up_tokenization_spaces behavior while the other is BOS/EOS insertion for a specific tokenizer." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Backend dependency packaging and leftover dummy classes are unrelated." + }, + { + "left": "issue:43479", + "right": "issue:46829", + "accept": false, + "reason": "Phi4Multimodal config initialization and VideoPrism output field wiring are different model-specific issues." + }, + { + "left": "issue:44568", + "right": "issue:47085", + "accept": false, + "reason": "A tokenizer special-tokens regression and a tokenizer thread-safety regression are different bugs." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor save_pretrained missing files and Qwen3.5 MoE from_pretrained errors are unrelated." + }, + { + "left": "issue:42673", + "right": "issue:46693", + "accept": false, + "reason": "Both are performance-related, but VRAM leakage in dataloader threads is not the same as FA2 long-context generation slowdown." + }, + { + "left": "issue:44186", + "right": "issue:47085", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crashes are separate from fast tokenizer thread-safety regressions." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Gemma4 attention-mask construction and MaskGenerationPipeline final-batch handling are unrelated." + }, + { + "left": "issue:45600", + "right": "issue:47059", + "accept": false, + "reason": "A broken auto_mappings reference and a huggingface_hub CLI compatibility failure are unrelated CI issues." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but the affected behavior and model scope differ; one is cleanup-space behavior, the other is wrong token IDs for Granite models." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer loading failure in v5 and BOS/EOS special-token insertion failure are different symptoms and likely different fixes." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Phi4Multimodal config defaults and Whisper word-timestamp decoding are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "A notebook syntax error and a documentation parameter-name mismatch are not the same issue." + }, + { + "left": "issue:39692", + "right": "issue:46909", + "accept": false, + "reason": "Both are broken documentation examples, but they target different models and failure modes." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Importing sentence-transformers fails with NameError, while the other is an ImportError for Qwen2.5-VL; unrelated imports." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but the reports describe different models and failure patterns rather than one shared concrete bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet. The Qwen3ForCausalLM pair is a clear duplicate with identical titles, and choosing issue:42673 as canonical is well-supported by the earlier timestamp and richer discussion trail. The remaining soft-edge rejections stay conservative and do not overstate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 49, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28282", + "issue:30333", + "issue:34689", + "issue:35141", + "issue:41628", + "issue:42222", + "issue:42673", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42898", + "issue:43116", + "issue:43295", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43874", + "issue:44060", + "issue:44075", + "issue:44079", + "issue:44190", + "issue:44222", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44448", + "issue:44485", + "issue:44623", + "issue:44734", + "issue:44843", + "issue:44933", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45200", + "issue:45245", + "issue:45310", + "issue:45356", + "issue:45405", + "issue:45656", + "issue:45712", + "issue:46132", + "issue:46231", + "issue:46762", + "issue:47085" + ], + "soft_pairs": [ + "issue:27957|issue:30333", + "issue:43479|issue:44079", + "issue:42673|issue:47085", + "issue:44843|issue:47085", + "issue:30333|issue:34689", + "issue:43874|issue:45092", + "issue:42898|issue:43295", + "issue:44295|issue:45656", + "issue:42222|issue:45200", + "issue:42898|issue:45356", + "issue:44734|issue:45103", + "issue:30333|issue:35141", + "issue:22355|issue:28282", + "issue:44222|issue:44297", + "issue:44075|issue:46231", + "issue:43116|issue:44190", + "issue:45405|issue:45712", + "issue:44297|issue:45656", + "issue:42831|issue:44623", + "issue:44933|issue:45712", + "issue:44448|issue:46132", + "issue:44222|issue:44623", + "issue:44060|issue:45125", + "issue:42796|issue:43723", + "issue:43479|issue:43644", + "issue:45245|issue:46132", + "issue:44373|issue:44485", + "issue:44360|issue:46762", + "issue:41628|issue:42757", + "issue:44297|issue:45310" + ], + "prepared_review_unit_hash": "2a8b63805be1d7732f9c7b68ed645c6a547d784c07b9d6119a8f525464433380", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36575, + "estimated_input_tokens": 9144, + "estimated_eval_tokens": 18544 + }, + "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; the set is a heterogeneous mix of tokenizer/import, multimodal, optimizer, serving, and training issues. If a representative issue is needed, issue 44843 is the strongest single example because it is a concrete, user-facing tokenizer loading regression with clear reproduction and broader impact.", + "confidence": 0.92, + "canonical_issue_reason": "issue:44843 is the best representative issue here: it describes a concrete core regression in AutoTokenizer loading, has clear failure mode (HF_HUB_OFFLINE), and is more broadly relevant than the more niche or model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44843 is the most suitable single issue to anchor the cluster, but only as a representative example\u2014not as a true duplicate of the rest.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27957", + "right": "issue:30333", + "accept": false, + "reason": "Different subsystems and failures: Flash Attention 2 model support vs MLflow reporting on failed training jobs." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Unrelated bugs: multimodal config initialization vs ModelOutput key assignment semantics." + }, + { + "left": "issue:42673", + "right": "issue:47085", + "accept": false, + "reason": "Both mention threading, but one is a Qwen3 VRAM leak in dataloader threads and the other is fast tokenizer thread safety; different code paths and bugs." + }, + { + "left": "issue:44843", + "right": "issue:47085", + "accept": false, + "reason": "Tokenizer offline-mode regression is not the same as Rust tokenizer thread-safety regression." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "MLflow job-status reporting and Llama 3.2 Vision model loading are unrelated issues." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Both are multimodal/model-loading bugs, but one is a missing image processor method and the other is a meta-initialization incompatibility; not the same underlying defect." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Both are v5 regressions around token/processor behavior, but they affect different APIs and code paths." + }, + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Position_ids buffer access error and deepspeed optimizer double-step are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Different model families and different multimodal setup bugs; not the same change." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but one is clean_up_tokenization_spaces behavior and the other is Kimi-K2.5 codec/regex handling." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Serving KV-cache indexing crash and auto-docstring annotation crash are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "MLflow reporting vs token-embedding resize/post_init reinitialization are different bugs." + }, + { + "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 for AutoModel; unrelated causes." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "FP8 save_pretrained behavior and tokenizer_config class naming are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument handling and docstring parameter mismatch are not the same bug." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Example-script failures in different scripts with different root causes; too broad to merge." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "PEFT version pinning and leftover dummy classes are unrelated maintenance issues." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer config serialization and deepspeed optimizer stepping are unrelated." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FineGrainedFP8 accuracy vs processor.save_pretrained missing files are unrelated." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Bad image_utils import and dummy class leakage are different export/import problems." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Pegasus v4/v5 output difference and AttentionInterface.register behavior change are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both involve saving, but one is FP8 moe save_pretrained and the other is processor save_pretrained file omission; different code paths." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Both involve Qwen3 variants, but one is a tied-weights warning bug and the other is missing tensor-parallel plan support." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Tokenizer BPE marker handling and AutoTokenizer.from_pretrained loading regression are different defects." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Multimodal config defaults and non-persistent buffer corruption are unrelated." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-count limit error and AttentionInterface.register behavior change are unrelated." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring typo/accuracy issue is not the same as GLM-5 RoPE implementation." + }, + { + "left": "issue:44360", + "right": "issue:46762", + "accept": false, + "reason": "Different algorithmic components with different correctness concerns; no shared code path." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Two distinct import errors from different packages and symbols." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer class serialization bug and Qwen3.5 MoE from_pretrained regression are different 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": 27, + "review_unit_index": 50, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:30990", + "issue:38175", + "issue:41628", + "issue:42831", + "issue:42907", + "issue:43122", + "issue:43232", + "issue:43475", + "issue:43504", + "issue:43723", + "issue:43824", + "issue:43986", + "issue:44038", + "issue:44190", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44295", + "issue:44360", + "issue:44373", + "issue:44484", + "issue:44623", + "issue:44829", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45310", + "issue:45357", + "issue:45834", + "issue:46002", + "issue:46050", + "issue:46082", + "issue:46209", + "issue:46305", + "issue:46328", + "issue:46369", + "issue:46424", + "issue:46431", + "issue:46614", + "issue:46655", + "issue:46726", + "issue:46909", + "issue:46932", + "issue:47059" + ], + "soft_pairs": [ + "issue:45071|issue:46082", + "issue:43723|issue:44977", + "issue:44222|issue:44484", + "issue:43504|issue:47059", + "issue:46614|issue:46932", + "issue:43122|issue:46209", + "issue:44263|issue:45092", + "issue:44295|issue:44623", + "issue:44263|issue:44373", + "issue:43232|issue:46424", + "issue:43232|issue:45834", + "issue:46209|issue:46431", + "issue:30990|issue:38175", + "issue:44190|issue:46082", + "issue:44263|issue:44360", + "issue:43475|issue:46305", + "issue:46305|issue:46655", + "issue:46369|issue:46909", + "issue:43986|issue:46726", + "issue:42831|issue:45310", + "issue:22355|issue:27957", + "issue:42907|issue:46050", + "issue:44279|issue:46328", + "issue:45092|issue:45357", + "issue:42831|issue:44038", + "issue:44261|issue:45092", + "issue:41628|issue:43824", + "issue:44829|issue:46002" + ], + "prepared_review_unit_hash": "45f997856c53e49b91f74c802445312eabf4d63bb260daff951170415479b1dd", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35169, + "estimated_input_tokens": 8793, + "estimated_eval_tokens": 17842 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated regressions and model-specific bugs across loading, save/load, quantization, generation, and processors. None of the soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45092 is the clearest, most self-contained report: it describes a concrete v5 meta-initialization incompatibility with old remote-code InternVL2 checkpoints and is more specific than the other candidates.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45092 is the strongest representative issue because it is specific, reproducible, and points to a concrete loading-path regression rather than a broad symptom or follow-on effect.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "Both are config-related, but one is a PretrainedConfig type-checking regression and the other is a LlamaConfig head_dim validation error; different code paths and failure modes." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading in v5 vs Qwen3.5 flash-attention generation failure are distinct problems affecting different subsystems." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "One is about FP8 save_pretrained for MoE models, the other about max_shard_size default behavior; both touch saving but not the same bug." + }, + { + "left": "issue:43504", + "right": "issue:47059", + "accept": false, + "reason": "Legacy BEiT preset loading and a CLI incompatibility with huggingface_hub are unrelated." + }, + { + "left": "issue:46614", + "right": "issue:46932", + "accept": false, + "reason": "Stale phonemizer language state in Wav2Vec2PhonemeCTCTokenizer is unrelated to Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenization drift across versions and FP8 inference failure on B200 are different issues with no shared code-path evidence." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split return handling in a MoE indexer is unrelated to remote-code checkpoint/meta-init incompatibility." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Reading position_ids after buffer registration and processor.save_pretrained missing files are separate serialization/attribute problems." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "A runtime torch.split behavior bug is not the same as a wrong docstring about position_ids." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation kwargs syncing after sync_gpus and cache max_cache_length semantics are distinct generation/cache issues." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation-state update logic and Kosmos2.5 OCR index overflow are unrelated." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "FP8 inference on B200 and missing chat template handling are different failure modes in different pathways." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading stall and siglip2 zero-probabilities are different model/runtime bugs." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading in an example script is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "A torch.split return-value issue and a missing ReLU in the DSA indexer are not the same bug." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "Missing fpn_position_embeddings in SAM3 video output and a sliding-window off-by-one are different SAM/video bugs." + }, + { + "left": "issue:46305", + "right": "issue:46655", + "accept": false, + "reason": "Both affect video processing, but one is a sliding-window mask off-by-one and the other is a padding calculation error; not the same defect." + }, + { + "left": "issue:46369", + "right": "issue:46909", + "accept": false, + "reason": "Memory overflow in sam3-litetext and a broken DIA documentation example are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:46726", + "accept": false, + "reason": "Both mention AutoProcessor/backend availability, but one is a crash for video models without torchvision and the other is an unrecognized image processor error; not enough to be the same underlying bug." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy problems and Qwen3.5 from_pretrained errors are separate issues." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Missing transformers.onnx module import and XLMRoberta Flash Attention 2 behavior are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Dequantized Ministrals/Devstrals save failures and sinq save/load breakage are both quantization-related but refer to different mechanisms and failure modes." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A general dependency issue is not the same as SequenceFeatureExtractor.pad converting arrays inefficiently." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Both involve Qwen3.5-family checkpoints, but one is load-time incompatibility with meta initialization while the other is save_pretrained writing incorrect visual encoder keys." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe from_pretrained/import behavior are different bugs." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm causing precision error is a different model/config issue than InternVL2 v5 meta-init incompatibility." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Two import errors for different symbols/models are not evidence of the same underlying bug." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Degenerate training with flash_attention_3 and NaNs in classification heads after checkout/init are distinct training stability 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": 27, + "review_unit_index": 51, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28218", + "issue:30990", + "issue:33357", + "issue:39379", + "issue:39401", + "issue:42371", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43540", + "issue:43576", + "issue:43644", + "issue:43646", + "issue:43784", + "issue:43976", + "issue:43986", + "issue:44188", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44327", + "issue:44373", + "issue:44483", + "issue:44568", + "issue:44945", + "issue:44964", + "issue:45278", + "issue:45381", + "issue:45406", + "issue:45593", + "issue:45753", + "issue:46002", + "issue:46172", + "issue:46305", + "issue:46369", + "issue:46402", + "issue:46909", + "issue:47059", + "issue:47217" + ], + "soft_pairs": [ + "issue:43408|issue:46402", + "issue:44964|issue:47059", + "issue:45381|issue:46305", + "issue:39379|issue:46172", + "issue:44483|issue:45406", + "issue:42831|issue:45753", + "issue:22355|issue:28218", + "issue:43644|issue:46305", + "issue:44327|issue:47059", + "issue:43986|issue:47217", + "issue:44188|issue:44373", + "issue:39401|issue:44568", + "issue:42757|issue:45278", + "issue:43408|issue:46369", + "issue:42886|issue:42994", + "issue:42886|issue:43066", + "issue:27957|issue:30990", + "issue:42796|issue:43122", + "issue:43232|issue:45593", + "issue:45381|issue:46909", + "issue:43540|issue:46909", + "issue:43646|issue:46002", + "issue:44295|issue:44297", + "issue:42907|issue:44945", + "issue:30990|issue:33357", + "issue:43576|issue:43784", + "issue:42907|issue:44964", + "issue:43976|issue:44945", + "issue:42371|issue:44263" + ], + "prepared_review_unit_hash": "3d1b2211c8a4ab2a7ed7e9297be5c42a2fabcbeaf3a2a69501ef7e799a3b9438", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 35803, + "estimated_input_tokens": 8951, + "estimated_eval_tokens": 18158 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of unrelated Transformers issues, mostly tokenizer/model regressions and version-specific breakages; there are no convincing duplicate pairs among the soft edges.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45278 is the broadest representative here: it describes a general post-upgrade regression with many import errors, which best matches the cluster\u2019s overall theme of Transformers version breakage. The rest are too specific and heterogeneous to serve as a canonical duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best single issue representative because it is the most general and centrally framed around upgrade-related breakage, while the others are narrower and mostly unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "Both involve SAM3, but one is a model-type warning during loading and the other is missing text-encoder weights; different concrete failures." + }, + { + "left": "issue:44964", + "right": "issue:47059", + "accept": false, + "reason": "One is a multimodal model load failure, the other is a Hugging Face Hub CLI attribute error; unrelated code paths." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Both mention masking/position behavior, but they are different models and different bugs: Qwen2.5-VL video positions vs Moonshine streaming window off-by-one." + }, + { + "left": "issue:39379", + "right": "issue:46172", + "accept": false, + "reason": "mllama mask handling and deepseek-v4 tensor transpose are unrelated issues." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "A chat/completions request validation issue is not the same as Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "FineGrainedFP8 accuracy regression and a Qwen3_5 image infinite loop are distinct bugs." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Module import failure for transformers.onnx and tokenizer extra-space behavior are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer junk values and Moonshine sliding-window masking are different code paths and symptoms." + }, + { + "left": "issue:44327", + "right": "issue:47059", + "accept": false, + "reason": "QA top-k bounds crash and CLI completion attribute error are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:47217", + "accept": false, + "reason": "Both are vision-language related, but one is a missing torchvision crash in AutoProcessor and the other is a placeholder-handling bug in VL models." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "A torch.compile attention-kernel divergence and a wrong docstring for position_ids are not the same bug." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer bugs, but they affect different models and different behaviors: offset mapping vs BOS/EOS insertion." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related regressions, but one is specifically a missing huggingface_hub symbol while the other is broad upgrade breakage; not the same underlying issue." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "Model-type mismatch warning and memory overflow in sam3-litetext are different problems." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Offline tokenizer cache loading and quantized model saving failure are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading and wrong tokenizer decoder type are different tokenizer/version issues, not one concrete bug." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta and Sentence-Transformers loading hang are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Both concern tokenization, but one is model-specific BPE marker handling and the other is version-to-version tokenization drift; not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation kwargs handling 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 bug and a broken DIA documentation example are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46909", + "accept": false, + "reason": "Qwen3OmniMoe video processing failure and a DIA docs issue are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:46002", + "accept": false, + "reason": "Buffer initialization junk values and NaNs in classification heads are different failure modes." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "A position_ids buffer access error and tokenizer_config class mismatch are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Quantized model saving failure and pipeline-parallelism output corruption are different code paths." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "A loading hang in Sentence-Transformers and a MacOS bus error in a CLIP model are not the same issue." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Both are v5-related breakages, but the CLI/env command regression and the sentence-transformers import NameError are distinct symptoms and subsystems." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Dequantized model saving and Phi-4 multimodal loading are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version compatibility and pipeline-parallelism output correctness are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API guidance and a torch.split return-value bug have no shared underlying cause." + } + ] + }, + "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": 27, + "review_unit_index": 52, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:29159", + "issue:36296", + "issue:39401", + "issue:41628", + "issue:42175", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43329", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:43937", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44222", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44488", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44829", + "issue:44938", + "issue:44945", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45290", + "issue:45478", + "issue:45600", + "issue:46172", + "issue:46326", + "issue:46614", + "issue:46762", + "issue:47059" + ], + "soft_pairs": [ + "issue:41628|issue:45278", + "issue:42907|issue:43937", + "issue:43756|issue:44060", + "issue:43784|issue:45600", + "issue:43329|issue:47059", + "issue:44117|issue:44704", + "issue:39401|issue:44488", + "issue:44625|issue:44829", + "issue:39401|issue:44521", + "issue:46172|issue:46762", + "issue:44360|issue:44485", + "issue:44704|issue:45290", + "issue:43408|issue:43475", + "issue:44038|issue:44297", + "issue:43122|issue:46326", + "issue:43122|issue:44521", + "issue:44038|issue:44295", + "issue:42907|issue:43452", + "issue:22355|issue:29159", + "issue:44038|issue:44222", + "issue:45216|issue:45478", + "issue:43824|issue:44351", + "issue:43122|issue:44779", + "issue:43116|issue:45200", + "issue:39401|issue:46614", + "issue:42757|issue:44351", + "issue:36296|issue:43116", + "issue:22355|issue:29127", + "issue:43278|issue:44945", + "issue:42175|issue:44938" + ], + "prepared_review_unit_hash": "c459ad399c091aa98604b45e439674c2ee4468f2494c29cf8bc4addc046c0789", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36548, + "estimated_input_tokens": 9137, + "estimated_eval_tokens": 18530 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like a true duplicate; the set is a loose mix of unrelated Transformers regressions with some broad import/tokenizer/version-similarity overlap. I\u2019d use the broad import-regression report as the best issue representative.", + "confidence": 0.83, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: multiple import failures after upgrading Transformers, which makes it the most representative issue for the cluster\u2019s import-regression subtheme.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most globally useful representative because it captures a wide class of breakages after upgrade, whereas the other issues are narrower and model- or API-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but one is AutoImageProcessor and the other is a broad post-upgrade import regression; not the same bug." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Different model paths and symptoms: save/dequantization failure vs invalid GenerationConfig." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Unrelated model issues: RoPE layer drop vs incorrect tied-weights warning." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Both affect imports, but one is a missing nn symbol and the other is stale auto_mappings config references." + }, + { + "left": "issue:43329", + "right": "issue:47059", + "accept": false, + "reason": "Completely different code paths: multimodal token counting vs CLI completion breakage from huggingface_hub." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Tokenizer mapping fallback bug is unrelated to AutoProcessor not forwarding kwargs to cached_file." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Qwen3 offset_mapping bug is not the same as a model-load failure for cjvt/sleng-bert." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "Config propagation bug for num_labels is unrelated to flash_attention_3 training degeneration." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Different issues in tokenization/chat templating versus offset mapping." + }, + { + "left": "issue:46172", + "right": "issue:46762", + "accept": false, + "reason": "Both mention numerical behavior, but they concern different models and different attention implementations." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "DSA indexer activation omission and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "Different chat/template bugs: missing kwargs forwarding vs assistant message crash with tool calls." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Same family, but warning about wrong model type is not the same as a missing fpn_position_embeddings attribute." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe load issue and tokenizer save metadata mismatch are different problems." + }, + { + "left": "issue:43122", + "right": "issue:46326", + "accept": false, + "reason": "Both are tokenizer-related regressions, but they affect different tokenizer implementations and failure modes." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Broad tokenizer regression is not the same as multimodal assistant mask handling." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "MoE/version issue and position_ids buffer-read error are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Different loading/saving code paths and different symptoms; not one underlying bug." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "ONNX module import error is unrelated to slow/fast tokenizer inconsistency." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Both involve MoE, but one is a model/version load issue and the other is FP8 save_pretrained behavior." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5 and recent versions, but one is checkpoint saving correctness and the other is from_pretrained loading error." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Distinct missing-symbol import errors for different APIs." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Two different tokenizer regressions with different models and outputs." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example-script classification output bug and Gemma 4 multimodal token-id requirement are unrelated." + }, + { + "left": "issue:39401", + "right": "issue:46614", + "accept": false, + "reason": "Different tokenizer/phonemizer subsystems and different state bugs." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing names from different packages/modules." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor parallel training bug and empty classification outputs are not the same issue." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "No module named transformers.onnx is unrelated to LayoutLMv3 error messaging." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype drift and incorrect LLM output under pipeline parallelism are different failures." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Backend dependency packaging issue is unrelated to Python 3.14 import/load failure." + } + ] + }, + "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": 27, + "review_unit_index": 53, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28348", + "issue:38617", + "issue:39379", + "issue:42371", + "issue:42673", + "issue:42898", + "issue:43066", + "issue:43116", + "issue:43232", + "issue:43493", + "issue:43611", + "issue:43618", + "issue:43825", + "issue:43874", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44188", + "issue:44355", + "issue:44483", + "issue:44485", + "issue:44610", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:45072", + "issue:45083", + "issue:45278", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45753", + "issue:46002", + "issue:46231", + "issue:46328", + "issue:46614", + "issue:46640", + "issue:46710", + "issue:46726", + "issue:46829", + "issue:46877", + "issue:46909", + "issue:46981", + "issue:47116", + "issue:47217" + ], + "soft_pairs": [ + "issue:39379|issue:43232", + "issue:43493|issue:45753", + "issue:44188|issue:46328", + "issue:46231|issue:46877", + "issue:42898|issue:46614", + "issue:43937|issue:44485", + "issue:38617|issue:45278", + "issue:44483|issue:47217", + "issue:43976|issue:45446", + "issue:44038|issue:45478", + "issue:43611|issue:46002", + "issue:43066|issue:44779", + "issue:42673|issue:45072", + "issue:45357|issue:45478", + "issue:44355|issue:46231", + "issue:44625|issue:46002", + "issue:43116|issue:43493", + "issue:45083|issue:45446", + "issue:44610|issue:46909", + "issue:42371|issue:43874", + "issue:44075|issue:44936", + "issue:43618|issue:46829", + "issue:46726|issue:46981", + "issue:46328|issue:47116", + "issue:28348|issue:46640", + "issue:43992|issue:44704", + "issue:43066|issue:46710", + "issue:43116|issue:43825" + ], + "prepared_review_unit_hash": "03398c13ef429135d5bdac557a04b2ab6d3f13ec0d4e3cf374b7b32bc0d477f8", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35754, + "estimated_input_tokens": 8939, + "estimated_eval_tokens": 18134 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft-edge pair looks like a true duplicate or the same fix; most are broad-theme similarities only (e.g. v5 regressions, AutoProcessor, Qwen3, flash-attention) rather than the same concrete bug. I\u2019d use the broad import-regression report as the best canonical issue in this set.", + "confidence": 0.89, + "canonical_issue_reason": "issue:45278 is the strongest umbrella-style bug report here: it describes a broad, user-visible regression after upgrading Transformers and can subsume related import-breakage reports better than the narrower model-specific tickets.", + "canonical_pr_reason": null, + "best_issue_reason": "Best representative issue because it is broad, clearly regression-oriented, and most suitable as a top-level reference point for related version-breakage reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39379", + "right": "issue:43232", + "accept": false, + "reason": "Different subsystems and failure modes: mllama masking logic vs generation kwargs after sync_gpus." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen/SigLIP-family models but describe different bugs: cross-implementation discrepancy vs an infinite loop on a specific image." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "Attention-kernel branching under torch.compile is unrelated to SequenceFeatureExtractor.pad conversion overhead." + }, + { + "left": "issue:46231", + "right": "issue:46877", + "accept": false, + "reason": "Documentation parameter-name mismatch is not the same as auto_docstring printing diagnostics at import time." + }, + { + "left": "issue:42898", + "right": "issue:46614", + "accept": false, + "reason": "Tokenizer cleanup behavior in v5 and a stale phonemizer backend are separate tokenization/phonemization issues." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GenerationConfig validation and GLM-5 RoPE implementation are different concrete problems." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Related release-breakage theme, but one is a specific missing symbol import and the other is a broad multi-import regression; not enough evidence they are the same underlying bug." + }, + { + "left": "issue:44483", + "right": "issue:47217", + "accept": false, + "reason": "Chat/completions request handling and VL image-placeholder handling are not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python-version compatibility issues do not match a flex_attention PyTorch version guard bug." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3 MoE, but one is a Qwen3-VL-Moe issue and the other is a Qwen3.5 MoE from_pretrained error; different code paths and model variants." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Model-loading base_model_prefix regression and NaNs in classification heads are unrelated." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "The tokenizer-decoder-type report may be a possible root cause, but the DeepSeek issue is not clearly the same concrete bug without more evidence." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leakage in Qwen3ForCausalLM under dataloader threads is unrelated to dtype mismatches in inference tests." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5 MoE, but save_pretrained key serialization and from_pretrained loading errors are different failures." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Compiled Python file execution errors are unrelated to a training API doc typo." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 config and NaNs in classification heads are different bugs." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script multi-label empty output and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Feature-length helper behavior in qwen3_omni_moe is unrelated to a flex_attention import/version check." + }, + { + "left": "issue:44610", + "right": "issue:46909", + "accept": false, + "reason": "Processor output size mismatch and a broken documentation example are not the same issue." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance and a missing image-patch-count method are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling is unrelated to trainer.evaluate() failing after train()." + }, + { + "left": "issue:43618", + "right": "issue:46829", + "accept": false, + "reason": "Both are output-field assignment issues, but they affect different model classes and different output members; not clearly the same fix." + }, + { + "left": "issue:46726", + "right": "issue:46981", + "accept": false, + "reason": "AutoProcessor backend-missing messaging and HTTP-429 retry behavior are distinct from each other." + }, + { + "left": "issue:46328", + "right": "issue:47116", + "accept": false, + "reason": "SequenceFeatureExtractor.pad performance overhead is unrelated to streaming pipeline regression." + }, + { + "left": "issue:28348", + "right": "issue:46640", + "accept": false, + "reason": "Flash-attention support request for GPT2LMHeadModel is not the same as an open Flash Attention 2 bug report." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "UMT5Encoder weight loading and AutoProcessor cached_file kwargs handling are different code paths." + }, + { + "left": "issue:43066", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer decoder-type regression and DeepSeek-R1 accuracy regression are too far apart to treat as the same concrete bug." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example classification script output and pipeline translation-task messaging 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": 27, + "review_unit_index": 54, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:33453", + "issue:39379", + "issue:39401", + "issue:42222", + "issue:42673", + "issue:42796", + "issue:43064", + "issue:43441", + "issue:43450", + "issue:43673", + "issue:43688", + "issue:43824", + "issue:43827", + "issue:43937", + "issue:43986", + "issue:44117", + "issue:44263", + "issue:44291", + "issue:44380", + "issue:44410", + "issue:44466", + "issue:44704", + "issue:44779", + "issue:44829", + "issue:44977", + "issue:44995", + "issue:45071", + "issue:45092", + "issue:45216", + "issue:45310", + "issue:45341", + "issue:45405", + "issue:45446", + "issue:45561", + "issue:45563", + "issue:45588", + "issue:46002", + "issue:46077", + "issue:46328", + "issue:46512", + "issue:46697", + "issue:46752", + "issue:46762", + "issue:46897", + "issue:46981", + "issue:47000", + "issue:47059", + "issue:47110" + ], + "soft_pairs": [ + "issue:43827|issue:45446", + "issue:42673|issue:42796", + "issue:44995|issue:46697", + "issue:45092|issue:46762", + "issue:33453|issue:39401", + "issue:44263|issue:46762", + "issue:43673|issue:46328", + "issue:43824|issue:46981", + "issue:43064|issue:43673", + "issue:39379|issue:43688", + "issue:42222|issue:43441", + "issue:42222|issue:47000", + "issue:45405|issue:45446", + "issue:43937|issue:45563", + "issue:44380|issue:46077", + "issue:44117|issue:45588", + "issue:42222|issue:46752", + "issue:44291|issue:45071", + "issue:44466|issue:46002", + "issue:39379|issue:46897", + "issue:44410|issue:45310", + "issue:39401|issue:44779", + "issue:43937|issue:44977", + "issue:45071|issue:47059", + "issue:43450|issue:43986", + "issue:45341|issue:46328", + "issue:44829|issue:45216", + "issue:44704|issue:47110", + "issue:45071|issue:46512", + "issue:45341|issue:45561" + ], + "prepared_review_unit_hash": "41b9cb991216a40709c102d1694a21e27ca3226c439901cc38cf72a9a76c7564", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36659, + "estimated_input_tokens": 9165, + "estimated_eval_tokens": 18586 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a heterogeneous set of unrelated Transformers bugs, so none of the soft-edge pairs look like true duplicates or mergeable PR-equivalents. Most pairs share only a broad subsystem (tokenizers, generation, processors, attention, training) but differ in the concrete code path and failure mode.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43824 is a high-signal, user-facing bug with multiple references and a clear concrete failure mode (importing a missing Qwen2_5_VL symbol), making it a good representative issue for this cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43824 is the strongest overall representative: specific, reproducible, externally visible, and broadly impactful compared with the many more niche or lower-signal items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() vs an incorrect PyTorch version gate for AuxRequest; different layers and unrelated fixes." + }, + { + "left": "issue:42673", + "right": "issue:42796", + "accept": false, + "reason": "Both are tokenizer/model bugs, but one is VRAM leakage in multithreaded loading and the other is BPE marker handling; different symptoms and code paths." + }, + { + "left": "issue:44995", + "right": "issue:46697", + "accept": false, + "reason": "GLM-Moe stale indexer cache crash vs Swin2SR over-padding; unrelated components and failures." + }, + { + "left": "issue:45092", + "right": "issue:46762", + "accept": false, + "reason": "Meta-initialization incompatibility for old remote-code checkpoints vs numerical mismatch in Minimax attention; not the same bug." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Both mention tokenizers, but loading regression and wrong offset mapping are different failures in different model families." + }, + { + "left": "issue:44263", + "right": "issue:46762", + "accept": false, + "reason": "torch.split return-value issue in GlmMoeDsaIndexer vs attention numerical differences; unrelated." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Generation cache missing during chunked prefill vs a padding performance inefficiency; not the same underlying bug." + }, + { + "left": "issue:43824", + "right": "issue:46981", + "accept": false, + "reason": "Missing Qwen2_5_VL import symbol vs repeated HTTP 429 retries in AutoProcessor; different failure mode and fix." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Optimizer-state corruption under FSDP2/PEFT loading vs GenerationMixin cache issue; different subsystems and root causes." + }, + { + "left": "issue:39379", + "right": "issue:43688", + "accept": false, + "reason": "mllama mask handling bug vs aux-loss normalization bug; both training-related but distinct logic." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Broken vitpose models vs Ministral FlashAttention failure; unrelated model families and code paths." + }, + { + "left": "issue:42222", + "right": "issue:47000", + "accept": false, + "reason": "vitpose breakage vs RTDetr loss infeasibility; no concrete overlap." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "PEFT minimum-version mismatch vs incorrect PyTorch version check for flex_attention; unrelated compatibility issues." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "Invalid GenerationConfig vs stale warning for num_return_sequences; different generation-validation bugs." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "GPT2 attention scaling ignored under SDPA/FlashAttention vs Gemma4 bidirectional mask construction; different model logic." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None vs flash_attention crash on s_aux=None; unrelated lookup vs runtime-attribute failures." + }, + { + "left": "issue:42222", + "right": "issue:46752", + "accept": false, + "reason": "vitpose model breakage vs apply_chat_template empty-list IndexError; no shared code path." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected _is_hf_initialized arg in init_empty_weights vs PretrainedConfig type-checking breakage; different API regressions." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "lm_head.weight serialization inconsistency vs NaNs in classification heads; distinct training/save behavior bugs." + }, + { + "left": "issue:39379", + "right": "issue:46897", + "accept": false, + "reason": "mllama masked-out mask handling vs Florence2 double-shift training-loss bug; different models and failure mechanisms." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Missing qwen3next projections in layer 0 vs Qwen3.5 MoE from_pretrained error; related family, but not the same concrete bug." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Qwen3 offset_mapping regression vs Deepseek tokenizer producing incorrect results; both tokenizer bugs but not the same issue." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Invalid GenerationConfig vs Qwen3.5 flash-attention generation failure; different bug classes." + }, + { + "left": "issue:45071", + "right": "issue:47059", + "accept": false, + "reason": "PretrainedConfig type checking regression vs cli failure with huggingface_hub 1.22.0; unrelated surfaces." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Batched video-processor output shape bug vs AutoProcessor crash without torchvision; both video-related, but different conditions and fixes." + }, + { + "left": "issue:45341", + "right": "issue:46328", + "accept": false, + "reason": "testing_utils bug vs SequenceFeatureExtractor.pad inefficiency; one is correctness/race-like, the other is performance." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "FlashAttention-3 causing degenerate training vs incorrect Qwen3.5 checkpoint saving; different code paths and symptoms." + }, + { + "left": "issue:44704", + "right": "issue:47110", + "accept": false, + "reason": "AutoProcessor not forwarding kwargs to cached_file vs special_tokens list lacking keys; unrelated input-shape/API bugs." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "Type-checking regression for PretrainedConfig vs stale config.dtype after from_config(dtype=...); separate config-handling issues." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "A bug in testing_utils vs xdist workers racing on captured_info.txt; both in testing utilities, but different failures and likely different 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": 27, + "review_unit_index": 55, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:43874", + "issue:43931", + "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:44877", + "issue:44933", + "issue:45071", + "issue:45081", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45478", + "issue:45632", + "issue:45656", + "issue:46178", + "issue:46209", + "issue:46231", + "issue:46298", + "issue:46617", + "issue:46752", + "issue:46762", + "issue:46894", + "issue:46981", + "issue:47116" + ], + "soft_pairs": [ + "issue:44117|issue:45081", + "issue:43408|issue:43742", + "issue:42617|issue:45341", + "issue:45632|issue:46981", + "issue:38617|issue:41628", + "issue:43122|issue:46298", + "issue:44877|issue:46231", + "issue:42222|issue:46178", + "issue:39401|issue:46209", + "issue:42371|issue:44261", + "issue:43937|issue:44496", + "issue:44484|issue:45310", + "issue:44261|issue:46762", + "issue:44933|issue:45278", + "issue:42371|issue:46617", + "issue:44410|issue:45478", + "issue:44075|issue:45656", + "issue:38617|issue:44220", + "issue:44263|issue:44485", + "issue:43066|issue:43122", + "issue:43208|issue:46894", + "issue:43097|issue:44493", + "issue:44077|issue:46981", + "issue:46752|issue:47116", + "issue:43673|issue:45341", + "issue:43931|issue:44410", + "issue:41628|issue:45071", + "issue:43874|issue:44261", + "issue:44484|issue:45478" + ], + "prepared_review_unit_hash": "c0d15bd4c505a13e6e78f648ecae8556863c7f50a63d464e8dec43e2d56c865b", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35345, + "estimated_input_tokens": 8837, + "estimated_eval_tokens": 17930 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose cluster of Transformers bug reports, mostly v5 regression and model/tokenizer-loading issues, but the soft pairs generally point to different code paths, symbols, or models rather than true duplicates.", + "confidence": 0.77, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report ('Many import errors after update from 4.57.0 to 5.5.0') and best matches the cluster's general upgrade-regression theme, though the overall set is quite heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most representative anchor because it is broad and generic enough to cover the shared v5 breakage theme; it is still only an imperfect canonical since most other items are narrower and unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-related, but one is a None mapping in from_pretrained and the other is a Mistral regex/backend_tokenizer crash; different code paths." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Different models and failures: sam3 type warning vs MobileLLM key error." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "Unrelated: a runtime failure in 3d_parallel.py vs a small testing_utils bug." + }, + { + "left": "issue:45632", + "right": "issue:46981", + "accept": false, + "reason": "Cache-path collision for trust_remote_code vs repeated 429 retries per config file; unrelated cache vs HTTP logic." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different symbols/modules and likely different missing exports." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Tokenizer output change vs a noisy logging message; not the same bug." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Config-loading failure vs documentation parameter-name mismatch." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "Vitpose breakage vs tensorflow_text backend mapping error; different subsystems." + }, + { + "left": "issue:39401", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer offset_mapping bug vs FP8 inference failure on B200." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 API/settings discussion vs MLA layernorm epsilon precision issue." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Invalid GenerationConfig vs unrecognized model/config type; unrelated loading errors." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Save-pretrained shard sizing question vs Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:44261", + "right": "issue:46762", + "accept": false, + "reason": "Both involve numerical behavior, but one is a missing rms_norm_eps/config precision issue and the other is an attention reference mismatch; different code paths." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "Specific missing image_utils import vs broad post-upgrade import errors; too different to merge." + }, + { + "left": "issue:42371", + "right": "issue:46617", + "accept": false, + "reason": "TF32 control settings vs FileNotFoundError in trust_remote_code cache loading." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen issues: missing projections in qwen3next vs Qwen3.5 MoE load failure." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Both optimizer-related, but one ignores SGD args while the other double-calls optimizer.step with deepspeed." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for a missing configuration utility symbol vs an audio feature extraction bug." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "torch.split return handling vs GLM-5 RoPE implementation; unrelated failures." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is decoder type selection and the other is changed tokenization output; not the same fix." + }, + { + "left": "issue:43208", + "right": "issue:46894", + "accept": false, + "reason": "Different model families and symptoms: xLSTM training blockers vs CohereASR loss double-shift." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "tie_embeddings removal vs unexpected position-id key; unrelated API/config regressions." + }, + { + "left": "issue:44077", + "right": "issue:46981", + "accept": false, + "reason": "PatchTSMixer post_init policy vs repeated HTTP 429 retries; different concerns." + }, + { + "left": "issue:46752", + "right": "issue:47116", + "accept": false, + "reason": "Empty-list chat template crash vs streaming pipeline regression; different code paths." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache issue during chunked prefill vs a testing_utils bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch vs qwen3next missing projections; different model-loading problems." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor import error vs PreTrainedConfig type-checking issue; not the same regression." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image-processor method vs MLA layernorm epsilon precision mismatch." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Shard-size question vs the Qwen3.5 MoE from_pretrained error." + } + ] + }, + "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": 27, + "review_unit_index": 56, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:27957", + "issue:38617", + "issue:42757", + "issue:42886", + "issue:42890", + "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:44351", + "issue:44360", + "issue:44410", + "issue:44483", + "issue:44485", + "issue:44829", + "issue:44987", + "issue:45003", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45440", + "issue:45464", + "issue:45588", + "issue:45753", + "issue:46050", + "issue:46123", + "issue:46172", + "issue:46298", + "issue:46369", + "issue:46655", + "issue:46664", + "issue:46752", + "issue:46897", + "issue:46981" + ], + "soft_pairs": [ + "issue:27957|issue:43352", + "issue:44483|issue:45464", + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:43441|issue:44829", + "issue:43931|issue:44038", + "issue:43012|issue:43577", + "issue:43097|issue:43452", + "issue:44261|issue:44485", + "issue:43986|issue:45341", + "issue:45003|issue:45341", + "issue:42886|issue:46981", + "issue:43531|issue:45341", + "issue:45753|issue:46369", + "issue:44117|issue:44987", + "issue:43874|issue:44263", + "issue:43097|issue:45588", + "issue:43673|issue:46123", + "issue:45440|issue:46172", + "issue:44410|issue:45375", + "issue:44038|issue:44410", + "issue:38617|issue:42757", + "issue:43097|issue:46050", + "issue:43064|issue:44315", + "issue:43986|issue:45372", + "issue:43208|issue:46897", + "issue:46655|issue:46664", + "issue:38617|issue:44351", + "issue:42890|issue:46298", + "issue:46298|issue:46752" + ], + "prepared_review_unit_hash": "522bc950efced0864059a2d784ddf60d28c91ebc698017181650b8ab5141a8ee", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36561, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are unrelated Transformers regressions spanning model-specific loading, flash-attention, generation, video processors, and utility/import issues. None of the soft pairs look like true duplicates of the same underlying bug, so I would not merge any of them as duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "If a single anchor is needed, issue 43097 is the broadest core API regression with downstream impact and inbound references, but the cluster is not a true duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43097 is the strongest representative issue because it describes a central user-facing regression in core Transformers behavior, not just a model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "Both mention Flash Attention 2, but they are different model-specific support failures with different code paths and fixes." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both concern chat/completions, but one is a general request-acceptance regression while the other is a streaming inference failure on Qwen3.5." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Completely different bugs: a torch.split return-value issue versus a DeepSeek v4 dtype/transpose problem." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "Unrelated defects in different model internals; shared DeepSeek-style ML context is not enough for duplication." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both involve flash attention, but one is a model load/runtime failure and the other is degenerate training with flash_attention_3." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Different Qwen3-VL-family failures with distinct symptoms: weight-shape mismatch versus a Transformers 5.0 compatibility issue." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Both are dtype-related, but one is a compile-time warning and the other is a model loading dtype propagation bug." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "A removed config/API issue is unrelated to gguf_file loading breakage in AutoTokenizer/AutoModel." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "Different model implementation bugs: missing rms_norm_eps in MLA versus a GLM-5 RoPE implementation issue." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "No overlap beyond being bugs in the repo; one is an AutoProcessor/torchvision dependency crash, the other is a testing_utils bug." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unrelated utility-layer issues; unsafe sys.modules access is not the same problem as the testing_utils bug." + }, + { + "left": "issue:42886", + "right": "issue:46981", + "accept": false, + "reason": "Both involve hub/network behavior, but offline cache loading and repeated 429 retries are different failures with different fixes." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "A Qwen3-MoE sliding_window bug is unrelated to a testing_utils issue." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "Different model families and symptoms: an infinite-loop image case versus a SAM3-LiteText memory overflow." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Both fail during loading, but one is a tokenizer mapping None case and the other is a physical-intelligence/fast load regression." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Different subsystems entirely: image processor token counting versus GlmMoeDsaIndexer torch.split handling." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "A config/API removal issue is unrelated to a flash_attention.py crash on sink-less models." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "Both touch generation-related code, but chunked_prefill cache handling and MaskGenerationPipeline batching are distinct bugs." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Both are DeepSeek-related, but they target different model versions and different concrete defects." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3next missing projections and Qwen3_5MoeVisionConfig missing a field are separate model/config issues." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Different Qwen-family regressions with different failure modes; the titles do not indicate the same underlying bug." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are ImportErrors, but they are for different missing symbols from different packages and likely separate dependency/version issues." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Config removal and quantized save/load breakage are unrelated regressions." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Different training stack problems: optimizer state corruption under FSDP2/PEFT versus Liger Kernel not being applied with model_init." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Both are loading-time crashes, but one is about missing torchvision for video processors and the other is a mistral_common import breakage." + }, + { + "left": "issue:43208", + "right": "issue:46897", + "accept": false, + "reason": "The xLSTM report bundles multiple training bugs, while the Florence2 issue is a specific double-shift loss bug; not the same defect." + }, + { + "left": "issue:46655", + "right": "issue:46664", + "accept": false, + "reason": "Both are SmolVLM video-processor issues, but incorrect padding calculation and num_frames semantics are distinct concrete bugs." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different modules; they are similar in shape but not the same bug." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "A flaky integration test from missing set_seed is unrelated to a noisy logging message." + }, + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "No meaningful overlap: log noise versus an IndexError in apply_chat_template on empty input." + } + ] + }, + "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": 27, + "review_unit_index": 57, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43502", + "issue:43550", + "issue:43825", + "issue:43827", + "issue:43873", + "issue:43906", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44062", + "issue:44077", + "issue:44117", + "issue:44162", + "issue:44222", + "issue:44368", + "issue:44380", + "issue:44451", + "issue:44484", + "issue:44655", + "issue:44704", + "issue:44734", + "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:46701", + "issue:46710", + "issue:46821", + "issue:47059", + "issue:47116" + ], + "soft_pairs": [ + "issue:44062|issue:44117", + "issue:43986|issue:44484", + "issue:43097|issue:43873", + "issue:43012|issue:46821", + "issue:44117|issue:46701", + "issue:46172|issue:46710", + "issue:43502|issue:43992", + "issue:44704|issue:45103", + "issue:44936|issue:46365", + "issue:44451|issue:47116", + "issue:44380|issue:46421", + "issue:44938|issue:46178", + "issue:43064|issue:47116", + "issue:42907|issue:45753", + "issue:42907|issue:44655", + "issue:43012|issue:43550", + "issue:45712|issue:46298", + "issue:43992|issue:44484", + "issue:43012|issue:44368", + "issue:43827|issue:47116", + "issue:43208|issue:43550", + "issue:43352|issue:43976", + "issue:43352|issue:44380", + "issue:43066|issue:43906", + "issue:43825|issue:44655", + "issue:44734|issue:47116", + "issue:43097|issue:46298", + "issue:45584|issue:46365", + "issue:44380|issue:46132", + "issue:44162|issue:46172", + "issue:44077|issue:46365", + "issue:44222|issue:45478", + "issue:42890|issue:47059" + ], + "prepared_review_unit_hash": "eb8cd46dba2d5a6d29552c8ec0c944e4a43a4176f89f56c447c71b8d9a659f47", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36820, + "estimated_input_tokens": 9205, + "estimated_eval_tokens": 18666 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous, with a loose subtheme of Transformers v5 regressions and compatibility issues, but no soft pair looks like a true duplicate or same underlying fix. I would anchor triage on the broad v5 streaming-regression report, while treating the rest as separate bugs.", + "confidence": 0.43, + "canonical_issue_reason": "Issue 47116 is the clearest broad anchor: it explicitly describes a Transformers v5 runtime regression in the streaming pipeline, which best matches the small subset of v5-related breakage reports here.", + "canonical_pr_reason": null, + "best_issue_reason": "47116 is the most suitable global triage anchor because it is broad, recent, and concrete; however the overall set is mostly heterogeneous and not a single duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "Different tokenizer-loading failures: AddedToken keyword collision vs TOKENIZER_MAPPING_NAMES returning None." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "Unrelated: missing torchvision during AutoProcessor loading vs a save_pretrained shard-size question." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Different bugs: a removed config/API in v5 vs quantization offloading behavior." + }, + { + "left": "issue:43012", + "right": "issue:46821", + "accept": false, + "reason": "Different models and failure modes: bf16 precision warning vs TimesFM window_size AttributeError." + }, + { + "left": "issue:44117", + "right": "issue:46701", + "accept": false, + "reason": "Tokenizer mapping null-handling vs an image-text pipeline failure are not the same code path." + }, + { + "left": "issue:46172", + "right": "issue:46710", + "accept": false, + "reason": "Different DeepSeek issues: one is a dtype/transpose bug, the other is an accuracy regression after upgrade." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Different loader problems: local_files_only network leak vs missing embed_tokens.weight in UMT5Encoder." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "Unrelated: cached_file kwargs passthrough vs docstring parsing crash with future annotations." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "Trainer evaluation after train is a runtime state bug; import failure due to datasets is a dependency/import issue." + }, + { + "left": "issue:44451", + "right": "issue:47116", + "accept": false, + "reason": "Different reports: model loading failure for ScandiBERT vs a streaming pipeline regression in v5." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "Different subsystems: GPT2 attention scaling under SDPA/FlashAttention vs cache lazy-init recompilation." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python version import failure and tensorflow_text backend mapping are unrelated." + }, + { + "left": "issue:43064", + "right": "issue:47116", + "accept": false, + "reason": "Different code paths: distributed optimizer-state corruption vs streaming pipeline regression." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Save/dequantization failure vs an image-generation infinite loop are not the same bug." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Both involve saving, but one is dequantized model saving and the other is Pipeline.save_pretrained; different objects and failure modes." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A PyTorch warning about bf16 precision is unrelated to a torch.compile SDPA failure." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Left is a repo hygiene/dummy-class issue; right is just a logging complaint." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing encoder weights on load is unrelated to the max_shard_size default question." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Different warnings with different causes: bf16 precision warning vs tie_word_embeddings config warning." + }, + { + "left": "issue:43827", + "right": "issue:47116", + "accept": false, + "reason": "Docs still referencing pipeline() is not the same underlying bug as a streaming pipeline regression." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba compile/SDPA failure are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "FlashAttention support in a model class is unrelated to Python version compatibility." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Both mention attention backends, but one is unsupported FlashAttention 2 and the other is ignored GPT2 scaling config." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Tokenizer decoder type mismatch is not clearly the same as the isolated reproduction of another issue." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline task-support messaging and Pipeline.save_pretrained are different problems." + }, + { + "left": "issue:44734", + "right": "issue:47116", + "accept": false, + "reason": "A serving KV-cache indexing crash is not the same as a streaming pipeline regression." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed v5 API behavior and a noisy log message are unrelated." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation bug is unrelated to the Trainer import problem." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT2 backend-specific attention scaling and AttentionInterface.register behavior changes are different issues." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage is unrelated to the DeepSeek dtype/transpose bug." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "patchtsmixer post_init validation and Trainer import failure are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both involve MoE/save/load, but one is FP8 save_pretrained and the other is a from_pretrained error; not the same concrete bug." + }, + { + "left": "issue:42890", + "right": "issue:47059", + "accept": false, + "reason": "Seed-related test flakiness and a cli completion API break 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": 27, + "review_unit_index": 58, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:29870", + "issue:36010", + "issue:36296", + "issue:36683", + "issue:38617", + "issue:42175", + "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:44514", + "issue:44625", + "issue:44655", + "issue:44683", + "issue:44938", + "issue:45216", + "issue:45276", + "issue:45479", + "issue:46050", + "issue:46077", + "issue:46082", + "issue:46365", + "issue:46877", + "issue:47059", + "issue:47110" + ], + "soft_pairs": [ + "issue:42757|issue:43824", + "issue:44380|issue:45479", + "issue:36683|issue:43582", + "issue:43064|issue:44380", + "issue:42175|issue:46365", + "issue:43012|issue:44683", + "issue:38617|issue:43502", + "issue:44483|issue:44514", + "issue:43756|issue:44485", + "issue:43618|issue:44380", + "issue:43618|issue:43761", + "issue:45276|issue:46077", + "issue:36010|issue:42757", + "issue:43208|issue:43526", + "issue:44038|issue:44230", + "issue:42994|issue:44655", + "issue:43299|issue:43867", + "issue:44625|issue:46365", + "issue:43352|issue:43937", + "issue:44625|issue:46077", + "issue:43611|issue:44655", + "issue:44355|issue:46877", + "issue:36683|issue:43475", + "issue:44655|issue:46050", + "issue:43976|issue:44938", + "issue:36683|issue:44077", + "issue:44938|issue:47059", + "issue:43867|issue:45216", + "issue:36683|issue:47110", + "issue:46077|issue:46082", + "issue:43618|issue:45216", + "issue:36683|issue:43525", + "issue:29870|issue:36296" + ], + "prepared_review_unit_hash": "df625cd503d269ca5cf19663b6a81ee1f6d238c4b27255404398ee05158eb224", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37387, + "estimated_input_tokens": 9347, + "estimated_eval_tokens": 18950 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab-bag of unrelated issues. The only convincing duplicate-looking pair is the two CLIP v5 regression reports; the rest differ in model, subsystem, or failure mode too much to merge safely.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 36683 is the strongest standalone bug report here: it is concrete, open, and highly discussed, with a clear model/config regression and active triage signal.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 36683 is the best overall representative because it is specific, impactful, and has the richest discussion trail among the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they involve different symbols and different dependencies; not the same bug." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "One is about GPT-2 attention scaling under SDPA/FlashAttention, the other is a zero-loss classification regression; unrelated failures." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3 config vocab_size is unrelated to Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT-2 attention scaling config." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "One is backend packaging for TensorFlow, the other is a Trainer import problem tied to datasets; different root causes." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "PyTorch precision warning during compile and flex_attention compile failure are different issues." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Importing a removed utility and ignoring local_files_only are separate API/behavior bugs." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "A chat/completions API rejection and a processor batching crash are different code paths." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Both mention RoPE, but they concern different models and different bugs, not one shared defect." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIP output regression and GPT-2 attention scaling are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": true, + "reason": "Both describe the same CLIP v5 forward/output regression: optional outputs from CLIPVisionModel are not being populated correctly." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Both are Gemma4, but one is embedding resize propagation and the other is bidirectional-attention masking; different bugs." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Different missing imports from different packages; not the same underlying failure." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both mention Qwen3-VL-Moe, but one is a load/regression issue and the other is fp8 inference support; not mergeable as one fix." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model save corruption and Pipeline save_pretrained support are different save paths with different semantics." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Both are model-loading problems, but the reported failure modes and triggers differ." + }, + { + "left": "issue:44625", + "right": "issue:46365", + "accept": false, + "reason": "Qwen3.5 label propagation and Trainer import dependency conflicts are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Flash Attention support for Nemotron and GLM-5 GenerationConfig validation are different model/config issues." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Label propagation and attention-mask construction are unrelated behaviors." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Base-model-prefix loading regression and Pipeline save_pretrained are different mechanisms." + }, + { + "left": "issue:44355", + "right": "issue:46877", + "accept": false, + "reason": "Compiled Python file loading errors and auto_docstring import-time stdout diagnostics are different issues." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3 config vocab_size and SAM3 video encoder missing attribute are distinct model-specific bugs." + }, + { + "left": "issue:44655", + "right": "issue:46050", + "accept": false, + "reason": "General Pipeline serialization and sinq quantized model persistence are different save/load problems." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Python-version compatibility complaints, but one is about Transformers 5.1 support and the other about Python 3.14 loading; not the same issue." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3 vocab_size and patchtsmixer post_init validation are unrelated." + }, + { + "left": "issue:44938", + "right": "issue:47059", + "accept": false, + "reason": "Python 3.14 import/load failure and huggingface_hub CLI breakage are different compatibility regressions." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "Different checkpoint/model-loading failures with different triggers and symptoms." + }, + { + "left": "issue:36683", + "right": "issue:47110", + "accept": false, + "reason": "Gemma3 vocab_size and special_tokens list/dict handling are unrelated." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "One is Gemma4 attention masking, the other is LlamaConfig head_dim validation; different code paths." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIP output regression and Qwen3.5 save_pretrained checkpoint corruption are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3 config vocab_size and Llama4 pad_token_id are different config attribute issues." + }, + { + "left": "issue:29870", + "right": "issue:36296", + "accept": false, + "reason": "BART decoder_attention_mask assignment and tensor-parallel training are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays conservative. The chosen canonical issue is reasonably supported by the highest discussion activity, and the one accepted soft pair (43618/43761) is a plausible same-regression CLIP output bug. I did not see any soft-edge verdicts that overstate duplication beyond the 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": 27, + "review_unit_index": 59, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28282", + "issue:29870", + "issue:33453", + "issue:36296", + "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:44222", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44734", + "issue:44855", + "issue:44908", + "issue:44964", + "issue:45125", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45850", + "issue:46202", + "issue:46285", + "issue:46365", + "issue:46617", + "issue:46664", + "issue:46851", + "issue:47000" + ], + "soft_pairs": [ + "issue:43064|issue:44492", + "issue:43010|issue:45278", + "issue:44075|issue:46365", + "issue:43992|issue:45310", + "issue:44908|issue:46851", + "issue:29870|issue:33453", + "issue:44393|issue:47000", + "issue:43867|issue:46202", + "issue:44355|issue:44855", + "issue:36296|issue:43575", + "issue:43010|issue:45850", + "issue:43867|issue:45375", + "issue:41669|issue:44246", + "issue:42890|issue:45561", + "issue:43618|issue:44077", + "issue:43010|issue:43976", + "issue:43992|issue:45478", + "issue:43867|issue:45357", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:44360|issue:45440", + "issue:42890|issue:44964", + "issue:43502|issue:44062", + "issue:42890|issue:44112", + "issue:42898|issue:46664", + "issue:44222|issue:45125", + "issue:44734|issue:46617", + "issue:43723|issue:44556", + "issue:43502|issue:44485" + ], + "prepared_review_unit_hash": "e87c79f4c3c8db7638d1409efb68c5c8f9a0657e9c3d5e28047500d37bb2a969", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36006, + "estimated_input_tokens": 9002, + "estimated_eval_tokens": 18260 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are unrelated issues around v5 regressions, model loading, tensor parallelism, and test failures. Only one pair looks plausibly like the same duplicate import-performance bug.", + "confidence": 0.48, + "canonical_issue_reason": "Issue 41669 is the clearest, most actionable report and the only one that plausibly subsumes another item: an import-time regression tied to `import *` usage in model modules.", + "canonical_pr_reason": null, + "best_issue_reason": "41669 is the strongest representative because it points to a concrete cause and user-visible slowdown; the rest are largely separate bugs in different code paths.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "Wrong subsystem and symptom: optimizer-state/PEFT training bug vs a cache-strategy typo." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Both mention v5 regressions, but one is about cache update semantics and the other about import failures; not the same bug." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "Unrelated: SGD args not used vs importing Trainer fails when datasets is installed." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Different model families and different failures; no clear shared code-path." + }, + { + "left": "issue:44908", + "right": "issue:46851", + "accept": false, + "reason": "Scheduler kwargs bug vs variable-length position_ids bug; unrelated." + }, + { + "left": "issue:29870", + "right": "issue:33453", + "accept": false, + "reason": "Bart generate() decoder mask error vs tokenizer-loading regression; different code paths." + }, + { + "left": "issue:44393", + "right": "issue:47000", + "accept": false, + "reason": "Different models and failures: Qwen3-VL bbox output vs RTDetr loss infeasible cost matrix." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "Model state_dict loading error vs Gemma pipeline roundtrip/tokenization issue; not the same bug." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both are import-related, but one is about compiled Python files and the other is a torch.jit parsing/IndentationError on Python 3.13." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Both involve tensor parallelism, but one is a training bug and the other is an OOM while loading a specific Qwen model." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "One is a specific cache/no_grad API change; the other is a broad complaint about many models breaking." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "State-dict loading failure vs missing vision-config field; different underlying problems." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": true, + "reason": "Both are about `transformers` import taking too long; 41669 identifies the likely root cause (`import *` in model modules), and 44246 reports the same slowdown symptom." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Test flakiness from missing seeds vs xdist race on a captured-info file; not the same issue." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "Different model/API bugs: missing attentions assignment vs optional `post_init` deprecation/constraint." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update no_grad bug vs scheduler argument handling; unrelated." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Same broad area of model loading, but different model versions and no concrete shared failure mode." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "State_dict loading error vs incorrect visual encoder keys on save_pretrained; different directions and symptoms." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "PyTorch missing ImportError vs tokenizer-loading regression; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "Different CLIP-related structure regressions; one is about attentions, the other about `.text_model` layout divergence." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer discussion vs DeepSeekV3 implementation divergence; no shared concrete bug." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Seed-related test instability vs model loading failure for Phi-4 multimodal; unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Local-files-only API leakage vs AddedToken constructor error; different code paths." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Seed-related test flakiness vs stale device override test on CI; different issues." + }, + { + "left": "issue:42898", + "right": "issue:46664", + "accept": false, + "reason": "Tokenizer cleanup behavior change vs video processor `num_frames` semantics; unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained issue vs missing tensor-parallel plan for Qwen3_5MoE; different problems." + }, + { + "left": "issue:44734", + "right": "issue:46617", + "accept": false, + "reason": "KV-cache continuation crash in serve vs trust_remote_code cache FileNotFoundError; unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer.from_pretrained regression vs checkpoint reload incompatibility after upgrade; related area, but not the same concrete bug." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Local-files-only network leak vs RoPE implementation discussion; unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted soft-edge pair, issue:41669 with issue:44246, is well-supported by their shared import-slowdown symptom, and the rejections do not overclaim equivalence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 60, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "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:44961", + "issue:45071", + "issue:45125", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:45850", + "issue:46132", + "issue:46143", + "issue:46144", + "issue:46617", + "issue:46664", + "issue:46726", + "issue:47059", + "issue:47110" + ], + "soft_pairs": [ + "issue:43874|issue:46617", + "issue:43824|issue:46617", + "issue:45850|issue:47110", + "issue:43010|issue:43065", + "issue:43576|issue:46132", + "issue:42548|issue:44351", + "issue:45125|issue:45478", + "issue:44387|issue:44556", + "issue:30333|issue:33453", + "issue:44355|issue:44749", + "issue:45125|issue:45375", + "issue:30990|issue:31795", + "issue:43827|issue:44961", + "issue:43502|issue:44961", + "issue:43502|issue:44222", + "issue:46664|issue:46726", + "issue:43010|issue:43232", + "issue:30333|issue:36296", + "issue:43299|issue:44556", + "issue:29870|issue:31795", + "issue:33453|issue:34689", + "issue:44355|issue:45071", + "issue:42757|issue:43576", + "issue:43576|issue:45071", + "issue:43576|issue:47059", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:43576|issue:43824", + "issue:44315|issue:44360", + "issue:44485|issue:44961", + "issue:44351|issue:46617", + "issue:43352|issue:44961", + "issue:36296|issue:43901", + "issue:43475|issue:46617", + "issue:46143|issue:46144", + "issue:41628|issue:42548", + "issue:45375|issue:47110", + "issue:44222|issue:44230" + ], + "prepared_review_unit_hash": "1ac602f56978a792778915b3dab834bcd6075c8d4571750f1e7be1e892bc6c59", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38592, + "estimated_input_tokens": 9648, + "estimated_eval_tokens": 19552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is heterogeneous: most items are unrelated import, loading, CLI, docs, quantization, or model-specific regressions. I found no true duplicate pair among the soft edges, so the cluster should not be merged into a single canonical bug/change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue safely represents the whole set; issue:43576 is the broadest/highest-signal regression, but it still does not subsume the other unrelated reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43874", + "right": "issue:46617", + "accept": false, + "reason": "Different failures: GLM46V multimodal token counting vs trust_remote_code cache loading FileNotFoundError." + }, + { + "left": "issue:43824", + "right": "issue:46617", + "accept": false, + "reason": "Different import problem and model family; no shared concrete code path." + }, + { + "left": "issue:45850", + "right": "issue:47110", + "accept": false, + "reason": "Broad complaint about many v5 breakages vs a specific special_tokens structure error." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "Cache update no_grad concern vs Sam3PixelDecoder dummy Conv2d; unrelated fixes." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "CLI env command regression vs AttentionInterface.register behavior change; different bugs." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Different missing exports: PreTrainedModel vs HybridCache." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3.5 MoE tensor-parallel plan issue vs from_pretrained loading error; not the same concrete bug." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 quantization OOM vs checkpoint reload incompatibility after version upgrade." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow job reporting bug vs tokenizer loading regression." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled Python file errors vs performance slowdown after upgrade; unrelated." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Tensor-parallel plan missing vs config field dropped by strict validation; different model issues." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Loading hang vs confusing forward-argument docs." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still reference pipeline() vs unrelated 'racoon' issue." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "local_files_only network leak vs unrelated 'racoon' issue." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Offline-mode network leak vs FP8 save_pretrained MoE bug." + }, + { + "left": "issue:46664", + "right": "issue:46726", + "accept": false, + "reason": "Different video-processor semantics vs missing-vision-backend AutoProcessor error." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Cache update decoration issue vs generation kwargs after sync_gpus." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLFlow reporting failure vs tensor parallel training bug." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Qwen3VL MoE load regression vs generic checkpoint reload incompatibility." + }, + { + "left": "issue:29870", + "right": "issue:31795", + "accept": false, + "reason": "BART generate() UnboundLocalError vs docs confusion about forward inputs." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer loading regression vs Llama 3.2 Vision model-loading failure; different affected paths." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file runtime errors vs PretrainedConfig type-checking regression." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing huggingface_hub symbol vs transformers env CLI regression." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "CLI env failure vs type-checking regression; unrelated." + }, + { + "left": "issue:43576", + "right": "issue:47059", + "accept": false, + "reason": "CLI env command regression vs tests/cli hub completion attribute error; not the same concrete bug." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Docs confusion vs missing layer_type_validation import error." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "Unrelated 'racoon' placeholder issue vs Gemma-4 audio positional encoding bug." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI env regression vs missing Qwen2_5_VL export; different issues." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger kernel application bug vs DSA indexer activation discussion; unrelated." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Different missing imports: AutoImageProcessor vs PreTrainedModel." + }, + { + "left": "issue:45375", + "right": "issue:47110", + "accept": false, + "reason": "Missing config field vs special_tokens list shape error; unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "FP8 save_pretrained bug vs FP8 inference support request; same broad area, different concrete change." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "kwargs propagation bug vs attention-interface usage bug in RoFormer; related theme but not the same underlying defect." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 27, + "review_unit_index": 61, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43867", + "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:44861", + "issue:44908", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45341", + "issue:45362", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46209", + "issue:46218", + "issue:46328", + "issue:46664", + "issue:47059", + "issue:47110", + "issue:47176" + ], + "soft_pairs": [ + "issue:44315|issue:46209", + "issue:43867|issue:47110", + "issue:43525|issue:47110", + "issue:44485|issue:46664", + "issue:45362|issue:45850", + "issue:43673|issue:44492", + "issue:44623|issue:45230", + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44360|issue:46664", + "issue:44273|issue:45003", + "issue:41669|issue:43673", + "issue:44393|issue:45071", + "issue:43576|issue:44556", + "issue:44393|issue:45753", + "issue:43525|issue:45850", + "issue:44230|issue:46095", + "issue:43572|issue:45850", + "issue:44556|issue:45092", + "issue:44492|issue:45850", + "issue:42371|issue:44492", + "issue:45850|issue:47059", + "issue:41669|issue:46218", + "issue:43408|issue:44315", + "issue:44246|issue:44273", + "issue:42890|issue:44908", + "issue:43576|issue:44908", + "issue:43576|issue:47176", + "issue:43352|issue:44393", + "issue:44295|issue:45230", + "issue:44297|issue:45230", + "issue:43976|issue:44393", + "issue:44908|issue:45071", + "issue:44492|issue:46328", + "issue:41669|issue:44351", + "issue:44704|issue:44908", + "issue:41669|issue:45341", + "issue:43976|issue:46218", + "issue:44077|issue:44861" + ], + "prepared_review_unit_hash": "58727041a14257a430335955bc40ed70af2a6c90f59115054eb41eb33b0ea17b", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 39, + "serialized_chars": 38175, + "estimated_input_tokens": 9544, + "estimated_eval_tokens": 19344 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose collection of unrelated transformer/model bug reports that only share superficial API/model-family similarity. I found no soft pair that clearly represents the same underlying bug or change, so all soft edges should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "No single strong canonical duplicate target stands out; issue #44273 is the closest umbrella candidate because it is open and broadly about lazy-loading/import behavior, but the overall cluster is too heterogeneous for a reliable canonical issue.", + "canonical_pr_reason": null, + "best_issue_reason": "#44273 is the best available issue to keep as a representative anchor only because it is open and relatively broad; however, it should not be treated as a true duplicate hub for the rest of this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Different problems: model_init/Liger-Kernel application versus FP8 inference failure on B200." + }, + { + "left": "issue:43867", + "right": "issue:47110", + "accept": false, + "reason": "One is a state_dict loading-order error; the other is a special_tokens type error. No shared code-path is evident." + }, + { + "left": "issue:43525", + "right": "issue:47110", + "accept": false, + "reason": "Llama4Config pad_token_id missing is unrelated to special_tokens list/dict handling." + }, + { + "left": "issue:44485", + "right": "issue:46664", + "accept": false, + "reason": "RoPE implementation for GLM-5 and video num_frames semantics are unrelated subsystems." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Specific chat crash on Qwen3.5-35B is not the same underlying bug as a vague 'many models break in v5' complaint." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression and a cache-strategy typo are different issues." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Missing processor save_pretrained files is not the same as an uninformative generic bug report." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading issues and TensorFlow backend detection are distinct failure modes." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Qwen3-VL-Moe bug versus missing tensor-parallel plan for Qwen3_5Moe; different concrete defects." + }, + { + "left": "issue:44360", + "right": "issue:46664", + "accept": false, + "reason": "DSA indexer activation discussion is unrelated to SmolVLM frame-count semantics." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "These are both import-related, but one is a generic lazy-loading complaint and the other is a specific sys.modules access issue; not enough to confirm same bug." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-time performance regression from import * usage is not the same as a chunked_prefill cache bug." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL bbox hallucination/error and PretrainedConfig type-checking breakage are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken env command and checkpoint reload incompatibility are different versioning issues." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen3_5/VL, but one is bbox output corruption and the other is an infinite-loop image case; not the same bug." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "A missing config attribute is not the same as a broad report that many models break in v5." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5-Moe does not match a deepseekv4 weight-conversion question." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "StableLm pad_token_idx regression is a specific config issue, not the generic v5 breakage report." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are v5 compatibility complaints, but one is checkpoint reload compatibility and the other is remote-code meta-init incompatibility; different defects." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "A cache-strategy typo is not the same as a broad 'many models break' report." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API guidance and cache-strategy naming are unrelated." + }, + { + "left": "issue:45850", + "right": "issue:47059", + "accept": false, + "reason": "General model breakage in v5 is unrelated to CLI completion errors caused by huggingface_hub changes." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import-time slowdown from import * does not match a module import failure for Qwen3_5ForCausalLM." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Model-type mismatch warning and Liger-Kernel/model_init behavior are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Import slowness and lazy loading malfunction are adjacent but not the same concrete bug based on the titles alone." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Missing set_seed in tests is unrelated to inverse_sqrt scheduler kwargs handling." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken env command and scheduler kwargs omission are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:47176", + "accept": false, + "reason": "Environment command breakage is unrelated to a shell-command allowlist bypass security bug." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention 2.0 support limitation for Nemotron and Qwen3-VL bbox issues are different model bugs." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "position_ids buffer handling is unrelated to the generic bug report." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer_class mismatch in save_pretrained is unrelated to the generic bug report." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility and Qwen3-VL output hallucination are unrelated." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs handling is unrelated to PretrainedConfig type checking." + }, + { + "left": "issue:44492", + "right": "issue:46328", + "accept": false, + "reason": "Cache-strategy typo and SequenceFeatureExtractor.pad inefficiency are different code paths." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import * performance regression is unrelated to missing HybridCache import/export." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor cached_file kwargs forwarding is unrelated to inverse_sqrt scheduler kwargs." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import performance regression is unrelated to a small testing_utils bug." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python-version incompatibility and module import failure are different issues." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "post_init policy for patchtsmixer and _tied_weights_keys list handling 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": 27, + "review_unit_index": 62, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:28348", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:43066", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43576", + "issue:43704", + "issue:43976", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45230", + "issue:45250", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46172", + "issue:46218", + "issue:46655", + "issue:46894", + "issue:47059", + "issue:47110", + "issue:47176", + "issue:47177" + ], + "soft_pairs": [ + "issue:31795|issue:33453", + "issue:43976|issue:47176", + "issue:44485|issue:47176", + "issue:43576|issue:44273", + "issue:44360|issue:45468", + "issue:43519|issue:46655", + "issue:44315|issue:44829", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:44485|issue:47177", + "issue:43519|issue:44485", + "issue:45468|issue:46894", + "issue:44655|issue:44861", + "issue:43976|issue:47177", + "issue:43576|issue:46218", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:45850|issue:47177", + "issue:45850|issue:47176", + "issue:44485|issue:45468", + "issue:43704|issue:44485", + "issue:46218|issue:47177", + "issue:46018|issue:46172", + "issue:43502|issue:43519", + "issue:42371|issue:43704", + "issue:45850|issue:46218", + "issue:43519|issue:44623", + "issue:44263|issue:45468", + "issue:44261|issue:45468", + "issue:43298|issue:46095", + "issue:43519|issue:43704", + "issue:43352|issue:46218", + "issue:44351|issue:46218", + "issue:44861|issue:47110", + "issue:42913|issue:43066", + "issue:44861|issue:47059", + "issue:28348|issue:45250", + "issue:44908|issue:44998", + "issue:44998|issue:47177", + "issue:44998|issue:47176", + "issue:44492|issue:44998", + "issue:44998|issue:45850", + "issue:44861|issue:45290", + "issue:31515|issue:38617", + "issue:44230|issue:44623", + "issue:28348|issue:43352" + ], + "prepared_review_unit_hash": "7a6bfce8ba2b99eae15f6c41210a40499eec99e940958a31467aa4dc5277ff85", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 47, + "serialized_chars": 40947, + "estimated_input_tokens": 10237, + "estimated_eval_tokens": 20730 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous: most items are standalone Transformers bugs, regressions, or feature requests with only loose topical overlap (loading, tokenizer, model behavior, cache, security, etc.). No soft-edge pair looks like a true duplicate or a single fixable code-path conflict.", + "confidence": 0.86, + "canonical_issue_reason": "issue:33453 is the most broadly representative concrete bug report in this set (a model/tokenizer loading regression), so it is the least-bad anchor for a heterogeneous cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:33453 is the strongest self-contained issue to use as a cluster anchor, but only as a representative loading-regression report, not as a true umbrella duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Both touch loading/docs around model inputs, but one is documentation confusion and the other is a tokenizer-loading regression; not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:47176", + "accept": false, + "reason": "Python-version support breakage and shell-operator allowlist bypass are unrelated issues." + }, + { + "left": "issue:44485", + "right": "issue:47176", + "accept": false, + "reason": "GLM-5 RoPE implementation and command-substitution allowlist bypass are different code paths and different classes of bugs." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken `transformers env` CLI and lazy loading failure are unrelated behaviors." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "A DSA indexer ReLU discussion is not the same issue as Gemma-4 audio positional encoding." + }, + { + "left": "issue:43519", + "right": "issue:46655", + "accept": false, + "reason": "Qwen3VL timestamp math and SmolVLM video padding are both vision-related, but they are distinct processor bugs." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger Kernel not applied in `model_init` and flash_attention_3 degenerate training are separate training-path failures." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report versus a specific Qwen3.5 MoE `from_pretrained` regression; not a duplicate." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading malfunction and inverse_sqrt scheduler kwargs handling are unrelated." + }, + { + "left": "issue:44485", + "right": "issue:47177", + "accept": false, + "reason": "RoPE implementation bug and tar/zip slip RCE are completely different issues." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Qwen3VL timestamp handling and GLM-5 RoPE implementation are distinct model-specific bugs." + }, + { + "left": "issue:45468", + "right": "issue:46894", + "accept": false, + "reason": "Gemma-4 audio positional encoding and CohereASR training-loss double-shift are different model behaviors, not the same underlying bug." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Saving pipeline objects and `_get_tied_weight_keys` attribute errors are unrelated serialization/util bugs." + }, + { + "left": "issue:43976", + "right": "issue:47177", + "accept": false, + "reason": "Python compatibility breakage and archive path traversal/RCE are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env CLI and import failure for Qwen3_5ForCausalLM are different surfaces and different fixes." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp calculation and scheduler keyword handling are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star performance issue and VRAM leak in Qwen3ForCausalLM are not the same bug." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion about forward inputs and Llama 3.2 model-loading regression are different problems." + }, + { + "left": "issue:45850", + "right": "issue:47177", + "accept": false, + "reason": "Broad version-breakage complaint and a command/file-write security issue are unrelated." + }, + { + "left": "issue:45850", + "right": "issue:47176", + "accept": false, + "reason": "General regression complaint and shell-operator allowlist bypass are not the same change or bug." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Different model families and different math/encoding subsystems; not duplicates." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:46218", + "right": "issue:47177", + "accept": false, + "reason": "Import failure for Qwen3_5ForCausalLM and archive traversal/RCE are unrelated." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "Shared-expert gating in DeepSeek-V4 and a `comb.to(...).transpose(...)` bug are distinct DeepSeek issues." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Local-files-only network leakage and Qwen3VL timestamp calculation are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 API settings and Qwen3ForCausalLM VRAM leak are different subsystems." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "Broad version regression complaint and module import failure are not the same underlying issue." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp math and missing save_pretrained files are different processor/save issues." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return values in a DSA indexer and Gemma-4 audio positional encoding are unrelated." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "MLA layernorm epsilon precision bug and Gemma-4 audio positional encoding are different numerical issues." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching not supporting audio models and DeepSeek-V4 weight-conversion question are not the same bug." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Qwen3VL timestamp calculations and VRAM leaks in Qwen3ForCausalLM are separate model-specific failures." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "NemotronHForCausalLM Flash Attention 2 support and Qwen3_5ForCausalLM import failure are different issues." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "HybridCache import failure and Qwen3_5ForCausalLM import failure are both import-related but concern different symbols and fixes." + }, + { + "left": "issue:44861", + "right": "issue:47110", + "accept": false, + "reason": "Tied-weights helper crash and special_tokens list/dict handling are different attribute/type bugs." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Tokenizer behavior change across v4/v5 and wrong tokenizer decoder type are related to tokenizers, but the reported failures and fixes are not clearly the same." + }, + { + "left": "issue:44861", + "right": "issue:47059", + "accept": false, + "reason": "Tied-weights helper crash and CLI breakage with huggingface_hub are unrelated." + }, + { + "left": "issue:28348", + "right": "issue:45250", + "accept": false, + "reason": "GPT2 flash-attention support request and a generic Flash Attention 2.0 issue are too broad to treat as the same concrete change." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug and an unrelated placeholder issue title have no meaningful overlap." + }, + { + "left": "issue:44998", + "right": "issue:47177", + "accept": false, + "reason": "Placeholder/unemployment issue and archive traversal/RCE are unrelated." + }, + { + "left": "issue:44998", + "right": "issue:47176", + "accept": false, + "reason": "Placeholder/unemployment issue and shell-operator bypass are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache-strategy typo and placeholder/unemployment issue are unrelated." + }, + { + "left": "issue:44998", + "right": "issue:45850", + "accept": false, + "reason": "Placeholder/unemployment issue and broad version-breakage complaint are unrelated." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight key crash and chat-template tool-call crash are different code paths." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow checkpoint loading and missing `layer_type_validation` import are unrelated import/loading problems." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 support request for qwen3-vl/qwen3.5 moe and missing processor save files are different feature/bug reports." + }, + { + "left": "issue:28348", + "right": "issue:43352", + "accept": false, + "reason": "Both mention Flash Attention 2.0, but they target different models and different support gaps; not a single fixable duplicate." + } + ] + }, + "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": 27, + "review_unit_index": 63, + "review_unit_count": 63, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-517", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:31515", + "issue:31795", + "issue:36296", + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43298", + "issue:43668", + "issue:43717", + "issue:43813", + "issue:44031", + "issue:44230", + "issue:45310", + "issue:46018", + "issue:46144" + ], + "soft_pairs": [ + "issue:29870|issue:31515", + "issue:31515|issue:36296", + "issue:30333|issue:31515", + "issue:31515|issue:31795", + "issue:44230|issue:45310", + "issue:43298|issue:44230", + "issue:43717|issue:46144", + "issue:42936|issue:46018", + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43668|issue:43813", + "issue:43240|issue:43668", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "a73980f8c6fff4b30211f1b420f24abeedda712d84aeec25d235c7c9825a26f6", + "component_budget": { + "node_count": 517, + "item_count": 517, + "soft_pair_count": 1905, + "serialized_chars": 987735, + "estimated_input_tokens": 246934, + "estimated_eval_tokens": 494124 + }, + "budget": { + "node_count": 17, + "item_count": 17, + "soft_pair_count": 18, + "serialized_chars": 14949, + "estimated_input_tokens": 3738, + "estimated_eval_tokens": 7732 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated bug reports, docs issues, config/type-hint nits, and model-specific feature requests. The soft pairs look like loose topical similarity rather than true duplicates, so none should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 43240 is the strongest standalone bug report in the set: it is concrete, code-path oriented, and has the most external discussion/inbound references. It is still not a true umbrella for the rest of the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "43240 is the best representative issue by activity/impact signals, but the overall set is too unrelated for a single duplicate canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29870", + "right": "issue:31515", + "accept": false, + "reason": "One is a BART generation UnboundLocalError; the other is a slow from_pretrained load report. Different failures and code paths." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint loading slowness and tensor-parallel training bugs are unrelated symptoms, not the same underlying defect." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLflow reporting after failed jobs is unrelated to model loading performance." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Performance complaint about loading checkpoints vs a documentation confusion issue; not the same bug." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen3.5/MoE and fp8/from_pretrained, but one is an fp8 inference support request and the other is a version-specific load error; not clearly the same fix." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching audio support and fp8 Qwen VL/MoE inference support are different feature gaps." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init-weights behavior and RoFormer attention-interface usage are separate model implementation issues." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2Former ignore_value handling and DeepSeek-V4 shared-expert gating concern different models and different logic." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "A config type-hint typo is not the same as a tokenizer regex warning regression." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Mask2Former ignore_value usage and loss-kwargs propagation are different code paths and fixes." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss-function kwargs forwarding and RoFormer attention plumbing are unrelated changes." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Model loss/ignore_value behavior and tokenizer warning behavior do not share the same underlying bug." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "Different model families and different implementation defects; not mergeable as one issue." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2Former loss handling and a typo in quantization text are unrelated." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Incorrect type hint versus typo in an attribute path; these are separate nits." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Loss kwargs forwarding and a config annotation typo are not the same defect." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "A design question about RMS norm and a type-hint correctness issue are unrelated." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former ignore_value behavior and ModernBERT config typing are separate issues with no shared fix." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "03819896d0fefa47e0321e9b2942a9f4fa741d5de4c822095d5490bfedf30303", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues are about unrelated feature requests: token truncation behavior, ReduceLROnPlateau-like training behavior, and Wav2Vec2 word time offsets. They do not describe the same underlying bug or change, so none should be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Different areas and requests: tokenizer truncation direction vs Wav2Vec2 word boundary timestamps. No shared underlying bug/change." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Tokenizer truncation option request is unrelated to a ReduceLROnPlateau-like scheduling feature request." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "Training-scheduler functionality and Wav2Vec2 time-offset extraction are unrelated feature requests, not the same fix." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-13244-3", + "nodes": [ + "issue:13244", + "issue:15980", + "issue:16618" + ], + "soft_pairs": [ + "issue:13244|issue:16618", + "issue:15980|issue:16618" + ], + "prepared_review_unit_hash": "90715683bc6f6889882af4444cbcd4eb0263f75eb5cf7cfbbb79701201107945", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2479, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2479, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues are not duplicates of each other. Two are about model-loading failures, but they point to different models and root causes; the private-model auth message issue is clearly separate. Issue 16618 is the broadest user-facing failure and the best cluster representative.", + "confidence": 0.86, + "canonical_issue_reason": "#16618 is the most general and central user-facing failure in the set: a failed model load for a common model name, with the broadest applicability and strongest discussion activity.", + "canonical_pr_reason": null, + "best_issue_reason": "#16618 best represents the cluster because it is the broadest loading-failure report, while the others are narrower and point to different causes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Different underlying problems: Tapas tokenization mismatch versus failure to load bert-base-uncased. The titles do not indicate the same code path or fix." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Different root cause and symptom: a bad auth/error message when downloading a private model versus an 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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-26350-2", + "nodes": [ + "issue:26350", + "issue:33260" + ], + "soft_pairs": [ + "issue:26350|issue:33260" + ], + "prepared_review_unit_hash": "73f15ea417e2167142b5d8aa8796047069ff9d45d36023b6735428e0fef96bee", + "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 separate community feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They are only superficially similar and should not be treated as duplicates.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different underlying requests and code paths: Flash Attention 2 support vs GGUF support. They share a broad theme of expanding architecture support, but not the same 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": 5, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "4dc9f03ea018c810bb421baed53321d50ca8c39f3c27450d95cf848808c7f531", + "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": "The cluster splits into two clear groups: distinct language-specific translation requests, and a duplicate-like placeholder-template subgroup with identical titles and near-identical body sizes. The placeholder subgroup looks mergeable as duplicates, but I would not treat the language-specific requests as duplicates of those template reports. The two Thai-title issues form a separate exact-duplicate pair; the different Thai-titled issue remains separate.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43834 is the best representative of the repeated placeholder-title subgroup: it is the earliest item in the largest same-title cluster and most cleanly captures the malformed template pattern.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43834 is the strongest overall issue representative because it anchors the largest evident duplicate-like cluster, while the language-specific translation requests are still distinct from each other.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Turkish translation request vs placeholder-title report; the titles indicate different concrete issues." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Turkish translation request vs placeholder-title report; similar boilerplate is not enough to merge them." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Turkish language-specific request and placeholder-template report are not the same underlying issue." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Different requested languages (Turkish vs Romanian), so these are separate translation issues." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Different requested languages (Turkish vs Spanish), so these are distinct issues." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish translation request vs placeholder-template report; no evidence they target the same concrete task." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Spanish translation request vs placeholder-template report; only the template structure overlaps." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Spanish and Romanian translation requests are separate issues." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Spanish translation request vs placeholder-template report; not the same underlying issue." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian translation request vs placeholder-template report; similar formatting does not make them duplicates." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian translation request vs placeholder-template report; not enough evidence of the same issue." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian translation request and placeholder-template report are different concrete issues." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "Romanian doc-translation request and the Thai-titled report are unrelated concrete issues." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "Romanian doc-translation request and the Thai-titled report are unrelated concrete issues." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": true, + "reason": "Identical placeholder-style title and nearly identical body size; this looks like a repeated malformed-template submission." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": true, + "reason": "Identical placeholder-style title and nearly identical body size; strong duplicate signal." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Different title and different apparent subject; soft similarity alone is not enough to merge them." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "Different title and different apparent subject; these do not clearly describe the same issue." + }, + { + "left": "issue:43834", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai title/topic; there is no concrete evidence these are the same report." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": true, + "reason": "Same placeholder title and almost identical body size; they look like duplicate malformed-template reports." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-template report vs Thai-titled report; the titles do not indicate the same underlying issue." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-template report vs Thai-titled report; not enough evidence of duplication." + }, + { + "left": "issue:44365", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai title/topic; similarity is only superficial." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-template report vs Thai-titled report; no clear common concrete issue." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-template report vs Thai-titled report; not enough evidence to merge." + }, + { + "left": "issue:44370", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai title/topic; the similarity score alone is not sufficient." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Same exact Thai title and identical body length; this is a strong duplicate pair." + }, + { + "left": "issue:46222", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai title and no explicit linkage; high similarity alone does not prove duplication." + }, + { + "left": "issue:46226", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai title and no explicit linkage; not clearly the same concrete report." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The high-level separation between the language-specific translation requests and the placeholder-template reports is plausible, but the summary overstates the evidence for the placeholder group by calling them exact duplicates. The packet only gives titles, body lengths, and soft-similarity scores; it does not show the bodies themselves. Also, the pairwise verdicts are internally inconsistent: 43834\u219444365 and 44365\u219444370 are accepted, but 43834\u219444370 is rejected despite even stronger similarity. That makes the duplicate-subgroup claim too strong for the available evidence." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "f52eab939b7c2da6741174d092e5192e5147e3615e0f0f313b7bcaa814cd41d4", + "component_budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28620, + "estimated_input_tokens": 7155, + "estimated_eval_tokens": 14566 + }, + "budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28620, + "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 mostly a grab bag of unrelated model-addition requests across vision, audio, and NLP. There are naming/theme overlaps, but no pair looks like the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.87, + "canonical_issue_reason": "No single issue is a true duplicate target; if one issue must anchor the cluster, issue 36879 is the broadest open model-add request and sits near the center of the similarity graph, but it is still only a representative.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 36879 is the most generic and central open request in this set, but the cluster spans multiple unrelated model families, so it is not a definitive canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36718", + "right": "issue:47004", + "accept": false, + "reason": "RoMa vs LoMa are different requested keypoint components; related naming, not the same change." + }, + { + "left": "issue:36718", + "right": "issue:41115", + "accept": false, + "reason": "Different model families and tasks: keypoint matching vs MiniCPM3 architecture support." + }, + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "Both are model requests, but for different models with no shared concrete implementation target." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config support and Qwen3-TTS support are unrelated changes." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "Different model requests: EuroBert config vs RF-DETR model addition." + }, + { + "left": "issue:36817", + "right": "issue:42954", + "accept": false, + "reason": "EuroBert config and MiMo are distinct model additions." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are different vision model additions." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR and LFM2.5 Audio are different modalities and code paths." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR model support and weight-sparse transformer support are different changes." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR and Qwen3-ASR are unrelated model support requests." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR vision model support is not the same change as Qwen3-TTS support." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different model families and tasks." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT addition and sparse-transformer support are separate implementation efforts." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio target different architectures and modalities." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are unrelated model additions." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are different model requests, not duplicates." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are unrelated." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 and weight-sparse transformer support do not describe the same concrete fix." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 architecture and ModernVBERT are distinct model additions." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 and Voxtral TTS support are different model integration tasks." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are unrelated 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 and Qwen3-TTS are not the same change." + }, + { + "left": "issue:42503", + "right": "issue:42954", + "accept": false, + "reason": "ModernVBERT and MiMo are separate model additions." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like RoPE support and Voxtral TTS support are unrelated requests." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT and BERT-like RoPE models are related only broadly, not as the same concrete change." + }, + { + "left": "issue:42738", + "right": "issue:42954", + "accept": false, + "reason": "BERT-like RoPE support and MiMo are different model requests." + }, + { + "left": "issue:42738", + "right": "issue:46653", + "accept": false, + "reason": "BERT-like RoPE support and Mirko Privitera are unrelated." + }, + { + "left": "issue:42954", + "right": "issue:44205", + "accept": false, + "reason": "MiMo and SAM3-LiteText are different model additions." + }, + { + "left": "issue:42954", + "right": "issue:43671", + "accept": false, + "reason": "MiMo and Qwen3-TTS are unrelated model support requests." + }, + { + "left": "issue:42954", + "right": "issue:45177", + "accept": false, + "reason": "MiMo and DeepSeek-OCR2 are different model requests." + }, + { + "left": "issue:42954", + "right": "issue:46653", + "accept": false, + "reason": "MiMo and Mirko Privitera are unrelated; one is a model request, the other is not." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are same vendor family but different concrete capabilities and code paths." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS and sparse-transformer support are not the same fix." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral TTS are both audio-related, but they target different models and implementations." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model requests." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio model integrations." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE are unrelated model additions." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support and sparse transformer support are different concrete changes." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are different audio models and tasks." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 are unrelated." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are different model requests." + }, + { + "left": "issue:43909", + "right": "issue:46645", + "accept": false, + "reason": "LFM2.5 Audio support and Granite SWA+sink option are not the same code-path change." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Sparse transformer support and EUPE model addition are unrelated requests." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are both speech-related, but they are distinct model integrations." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS are unrelated model requests." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 target different modalities and implementations." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are both vision-language related, but they are different model additions." + }, + { + "left": "issue:42503", + "right": "issue:46653", + "accept": false, + "reason": "ModernVBERT and Mirko Privitera are unrelated." + }, + { + "left": "issue:45027", + "right": "issue:46653", + "accept": false, + "reason": "Voxtral TTS support and Mirko Privitera are unrelated; no shared code change." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-38740-2", + "nodes": [ + "issue:38740", + "issue:46443" + ], + "soft_pairs": [ + "issue:38740|issue:46443" + ], + "prepared_review_unit_hash": "631a614f3578018bd0c27d8cda67f808a10e43abdd724f6abe2d5ffbacc34e81", + "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": "These two issues do not appear to describe the same underlying problem: one is a docs request about adding Pruna as an optimization framework, while the other is an open technical refactor for Vision RoPE. They should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No clear canonical issue for the pair: the topics, scopes, and expected fixes are unrelated (documentation/integration request vs. model architecture refactor).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #46443 is the better representative of the cluster only because it is a substantive code-related issue; #38740 is a documentation request and is not a duplicate partner for the refactor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Different subjects and remediation paths: a docs request to add Pruna versus a Vision RoPE refactor. No shared code path or concrete fix target is evident." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 8, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41084-2", + "nodes": [ + "issue:41084", + "issue:43037" + ], + "soft_pairs": [ + "issue:41084|issue:43037" + ], + "prepared_review_unit_hash": "ee37ce541e43b85acad762757e3a4ae2eaf5b5275e723f30f19722852bf64805", + "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": "These two issues are related only loosely by NLP/LLM inference context, but they describe different work: one is about block decoding mechanics, the other is a model-specific DeepSeek v3.2 support request. I would not triage them as duplicates.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "#43037 is the better standalone issue because it is a concrete feature request with a clear target (DeepSeek v3.2 support), while #41084 is more ambiguous and implementation-oriented.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Reject: the titles point to different underlying changes. \"Set Block Decoding\" sounds like an internal decoding-mode/algorithm setting, while \"DeepSeek v3.2 support\" is model compatibility work. There is no clear shared bug or code path, 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": "issue", + "component_index": 9, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41306-2", + "nodes": [ + "issue:41306", + "issue:41826" + ], + "soft_pairs": [ + "issue:41306|issue:41826" + ], + "prepared_review_unit_hash": "a5d29226d408e70c3d8fe504a842d4b0549dfc75a07047f44db7a63a958704d8", + "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 about different feature requests in different parts of the model stack: one for a TiledFusedLogitsLoss / Liger FLCE flag in `from_pretrained()`, the other for integrating TiledMLP to reduce memory footprint. They do not look like the same underlying bug or change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41306", + "right": "issue:41826", + "accept": false, + "reason": "Different requested capabilities and code paths; only broad memory/performance theme overlaps, which is not enough to treat them as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43598-2", + "nodes": [ + "issue:43598", + "issue:43600" + ], + "soft_pairs": [ + "issue:43598|issue:43600" + ], + "prepared_review_unit_hash": "14257f0c1f57a5d4c0a50095a85263bbed254dba8e62c4b10bbf88441bd2b6e7", + "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": "The two issues are related to custom trainers, but they describe different problems in different Trainer code paths: unused-column handling vs FLOPS estimation.", + "confidence": 0.87, + "canonical_issue_reason": "Issue #43598 is the more concrete and representative report, focused on a specific Trainer behavior (`remove_unused_column`) and its impact on customizability.", + "canonical_pr_reason": null, + "best_issue_reason": "#43598 is the better cluster representative because it names the specific mechanism being questioned and reads as the more actionable root concern.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "They are both about custom trainers, but the underlying bugs differ: one concerns `remove_unused_column` behavior, the other concerns how FLOPS are estimated for custom trainer classes. Different fixes/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": 11, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43684-2", + "nodes": [ + "issue:43684", + "issue:44405" + ], + "soft_pairs": [ + "issue:43684|issue:44405" + ], + "prepared_review_unit_hash": "50204f3d80fef2dc2638e429a08b9730b6a0c372e73f7bd06d87d3f461d92049", + "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 a broad transformer-automation level, but they request different model families and different AutoModel heads, so they should not be clustered as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #43684 is the broader, more foundational registration request: adding Qwen3-Omni to AutoModel and AutoModelForConditionalGeneration.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43684 is the better representative because it covers a core model registration gap rather than a narrower head-specific feature request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying requests: Qwen3-Omni model registration for AutoModel/AutoModelForConditionalGeneration vs Qwen3.5 AutoModelForSequenceClassification support. Same ecosystem, 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": 12, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "e31cbccf2e53bff336b25064158fe7cde62f2f4f0d09dba2b119668581ae51f7", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three issues share the same generic title \"huggingface \" and are likely duplicates of one another; \"Fett\" appears unrelated. I\u2019d keep issue 43845 as the canonical issue among the huggingface trio because it is the earliest and most representative entry.", + "confidence": 0.72, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface \" reports and is a clean representative for the duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "43845 is the best issue to anchor the cluster: earliest timestamp among the near-identical huggingface reports, with the other two likely dupes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles and no supporting detail tying \"Fett\" to the huggingface reports." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no evidence of the same underlying bug or request." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different titles; nothing indicates the same issue as the huggingface reports." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "Same title and highly similar sparse content; these look like duplicate reports of the same issue." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "Same title and clearly overlapping generic report; likely the same underlying issue." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Same title and near-identical high-level report; plausible duplicates of the same issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative: 43845 is the earliest of the repeated \"huggingface \" issues, 43835 has a different title and no supporting link, and the duplicate verdicts among 43845/43846/43859 are supported by the exact title match plus the soft-similarity evidence. No overstatement beyond the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44371-6", + "nodes": [ + "issue:44371", + "issue:44418", + "issue:44572", + "issue:44573", + "issue:44937", + "issue:47036" + ], + "soft_pairs": [ + "issue:44371|issue:44572", + "issue:44371|issue:44573", + "issue:44371|issue:44418", + "issue:44371|issue:47036", + "issue:44371|issue:44937", + "issue:44418|issue:44572", + "issue:44418|issue:44573", + "issue:44572|issue:44573", + "issue:44572|issue:47036", + "issue:44572|issue:44937", + "issue:44573|issue:47036", + "issue:44573|issue:44937" + ], + "prepared_review_unit_hash": "cac0291dbcb9a03e6ee019ea7f265a9c8c52d60c1d8c9afaccba26dd39215a2b", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7358, + "estimated_input_tokens": 1840, + "estimated_eval_tokens": 3936 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7358, + "estimated_input_tokens": 1840, + "estimated_eval_tokens": 3936 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are not duplicates of one another; the cluster looks like unrelated spam/generic placeholder issues rather than a single concrete bug or request.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items are unrelated low-signal/spam posts with no shared concrete problem to unify around.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a good global representative because none describes a real, specific underlying bug or enhancement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are generic spam placeholders, but there is no evidence they describe the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Same generic spam title is not enough to conclude a shared bug or request." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "One is spam and the other is a documentation suggestion; they are unrelated." + }, + { + "left": "issue:44371", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder versus community benchmark tip; no common concrete issue." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder versus a specific outreach-style title; not the same underlying problem." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "Documentation suggestion and spam placeholder are different kinds of posts with no shared bug." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "Documentation suggestion and spam placeholder are unrelated." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Identical spam titles alone do not establish the same underlying issue." + }, + { + "left": "issue:44572", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder and benchmark tip are unrelated." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder versus outreach-style title; no shared concrete issue." + }, + { + "left": "issue:44573", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder and benchmark tip are unrelated." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder versus outreach-style title; not a duplicate." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45145-2", + "nodes": [ + "issue:45145", + "issue:45151" + ], + "soft_pairs": [ + "issue:45145|issue:45151" + ], + "prepared_review_unit_hash": "067ad7d90c043812202e83ad80b475e398a0e1d95855d0159a3b1cea6bb78dfe", + "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 pair looks like a likely duplicate issue proposal: the titles are identical, the bodies are similarly sized, and there is a soft textual-similarity signal with no distinguishing metadata. #45145 is the better canonical issue because it was created earlier and already has the only inbound reference.", + "confidence": 0.86, + "canonical_issue_reason": "#45145 is earlier and has the only inbound reference, so it is the best canonical issue to keep for this likely duplicate pair.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the strongest representative issue: it appears to describe the same proposal as #45151 and has the slight advantage of earlier creation plus an inbound reference.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "Likely the same underlying proposal: identical titles, similar body length, and a soft similarity signal. The evidence is suggestive rather than conclusive, but it is enough to treat them as duplicates." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The canonical choice is reasonably grounded (earlier creation and the only inbound reference), but the duplicate verdict is a bit too strong for the packet. The evidence shown is identical titles, similar body lengths, and a soft-similarity signal; it does not directly establish that the bodies are the same or that there is no meaningful difference. The confidence of 0.99 is also too high for soft evidence only. A more conservative phrasing would be that they likely duplicate the same proposal." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45762-2", + "nodes": [ + "issue:45762", + "issue:45763" + ], + "soft_pairs": [ + "issue:45762|issue:45763" + ], + "prepared_review_unit_hash": "f29669185cace2791c0c7b6837b31c2c0abff251a92bb319610e907adf350fa0", + "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": "These two issues are unrelated empty posts with different personal names and dates; there is no evidence they describe the same bug, request, or change.", + "confidence": 0.99, + "canonical_issue_reason": "Issue #45762 is not meaningfully better than #45763; both are blank, non-technical reports, so there is no true canonical issue in this cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a good representative because neither contains actionable problem details or duplicate content.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "Different names and birth dates, with no shared problem statement or code/change context; the similarity is only superficial." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45764-2", + "nodes": [ + "issue:45764", + "issue:45765" + ], + "soft_pairs": [ + "issue:45764|issue:45765" + ], + "prepared_review_unit_hash": "a86ced27b18d47c2be66e2b72303541c71e79a30ec9dae59c75a9953c6eac662", + "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; they look like unrelated placeholder/person-name entries with different names and dates, and there is no shared technical context.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the items do not describe the same bug or request, so there is no duplicate target to consolidate around.", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good canonical representative; if one must be chosen, 45764 is only marginally cleaner formatted, but it is not meaningfully more canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "Different names and different dates with no shared problem statement or technical context; this is not 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": 18, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45810-2", + "nodes": [ + "issue:45810", + "issue:46295" + ], + "soft_pairs": [ + "issue:45810|issue:46295" + ], + "prepared_review_unit_hash": "e0b9647706b23ee76e0ba3a125723894cdbc8e037126f94f2b882ed0460e3ed8", + "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": "These two issues are unrelated feature requests: one asks for a new Qwen3_5 token-classification/value-model class, while the other requests FP8 training support. They should not be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45810 is the more specific request and reads like a concrete model-class addition, so it is the better representative of this cluster even though it is not a duplicate of 46295.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different underlying requests: model/task support for Qwen3_5 token classification vs training-time FP8 support. No shared code-path or same 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": 19, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "9ad8a8fe6beddd30c0dc8898ce7a116041874e61aebd10de13bf2c797d37af9d", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5723, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5723, + "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 one unrelated dataset issue plus two likely duplicate pairs: 45870/45871 and 45872/45873. The evidence is strongest on matching titles and matching body lengths for each pair, so the duplicate calls should stay conservative rather than claiming full-body identity.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 45870 is a good canonical representative for the 45870/45871 pair: it shares the same title and body length as 45871 and has slightly more discussion activity. It should not be treated as a representative for the separate 45872/45873 pair or the unrelated dataset issue.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45870 is the best single issue to keep as a representative item because it anchors one of the likely duplicate pairs and has marginally more engagement than 45871. The cluster still contains a second separate likely-duplicate pair and an unrelated issue, so no single issue fully covers the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Different topics: a dataset bundle listing versus a DataForge example request. Similar finance wording is not enough to make them the same issue." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Different topics: dataset bundle listing versus example request; no evidence they describe the same underlying issue." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Likely duplicate pair: same title and matching body length, with no signs of a different scope." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Related DataForge/financial wording, but the requests differ in scope: one is an example request and the other adds a scraping-service component." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Same broad theme as 45870/45872, but not the same concrete request." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Related wording only; these appear to be different requests rather than the same underlying issue." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Related wording only; the example request and the scraping-service idea do not look identical in scope." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Likely duplicate pair: same title and matching body length, with no evidence of a scope difference." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level conclusion is mostly grounded, but the analyst overstates the evidence in several places. From the packet we can see identical titles and matching body lengths for 45870/45871 and 45872/45873, but not the full bodies, so calling them \"exact duplicates\" or saying they have the \"same text\" is stronger than the packet supports. The summary should stay conservative (e.g., \"likely duplicate pairs\" or \"same title and same body length\"). The 45869-vs-others non-match is reasonable, but the canonical and soft-edge rationales need to avoid claiming exact identity that is not directly evidenced." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46439-2", + "nodes": [ + "issue:46439", + "issue:46441" + ], + "soft_pairs": [ + "issue:46439|issue:46441" + ], + "prepared_review_unit_hash": "b2dcfe5b1b28d9472aab7cf4a5816ab0d82ac79ccb766b6d687d853c102dfca2", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1640, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1640, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues appear related around static cache and `linear_attention`, but the packet only provides titles and a soft-similarity signal, so the exact relationship is unconfirmed.", + "confidence": 0.66, + "canonical_issue_reason": "Issue 46439 is the more general and diagnostic report, since it describes the broader static-cache/linear-attention failure in a way that can cover multiple symptoms.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46439 is the best representative issue because it is the more mechanism-oriented and less model-specific of the two.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46439", + "right": "issue:46441", + "accept": false, + "reason": "The titles suggest a possible connection around static cache and `linear_attention`, but the packet does not establish that they are the same underlying bug or the same code path." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The linkage is plausible, but the analyst overstates the evidence. The packet only gives titles plus a soft-similarity signal; it does not establish that these are the same underlying bug or the same code path. Phrases like \"same static-cache/linear-attention failure mode\" and \"same code path\" are stronger than the packet supports. A more conservative summary would say the two issues appear related around static cache and linear_attention, but the exact relationship is unconfirmed." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 21, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46551-2", + "nodes": [ + "issue:46551", + "issue:46552" + ], + "soft_pairs": [ + "issue:46551|issue:46552" + ], + "prepared_review_unit_hash": "c36bf37905a96077b5383f306cba80ca8b285a50af744daaa2b24c7b190cc49f", + "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 near-perfect duplicates: same title, same apparent body, and the pair is flagged as maximally similar. Issue #46552 is the better canonical thread because it is open and already has inbound references.", + "confidence": 0.99, + "canonical_issue_reason": "#46552 is open and already attracting references, so it is the better live tracker for this identical bug report.", + "canonical_pr_reason": null, + "best_issue_reason": "#46552 is the strongest issue to keep as the cluster representative: same report as #46551, but open and more likely to be the active discussion thread.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46551", + "right": "issue:46552", + "accept": true, + "reason": "Same title and same issue content; this is the same underlying bug report." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded in the packet: the issues share the same title, the pair has maximal soft-similarity (jaccard 1.0), and #46552 is open with inbound references, making it the better canonical thread." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 22, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46657-2", + "nodes": [ + "issue:46657", + "issue:46692" + ], + "soft_pairs": [ + "issue:46657|issue:46692" + ], + "prepared_review_unit_hash": "12746c6913e133b54fe8d0cf7a1e4130a26bb1690acb19eeda9d944b4ed63c8f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1569, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1569, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues describe different failure modes during training: one is a tensor-size mismatch, the other is an out-of-memory condition after several epochs. They are not the same underlying bug, so the soft duplicate link should be rejected.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 46657 is the cleaner representative if one must be chosen, because it states a concrete training error with a clear symptom; 46692 is a different resource-exhaustion problem, not the same defect.", + "canonical_pr_reason": null, + "best_issue_reason": "46657 is the best issue to anchor the cluster because it is a specific, actionable bug report; 46692 appears to be an unrelated training instability/OOM report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46657", + "right": "issue:46692", + "accept": false, + "reason": "Different symptoms and likely different root causes: tensor shape mismatch versus OOM after epochs. Shared training context is too broad to treat as a duplicate." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46822-2", + "nodes": [ + "issue:46822", + "issue:46823" + ], + "soft_pairs": [ + "issue:46822|issue:46823" + ], + "prepared_review_unit_hash": "2b7b5876fa34e0e0053c802631ea2d2f148dd825ffbeebffd0e284af1f69504a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1674, + "estimated_input_tokens": 419, + "estimated_eval_tokens": 1094 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1674, + "estimated_input_tokens": 419, + "estimated_eval_tokens": 1094 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues describe different bugs in different loading paths: ZeRO-3 zero.Init partitioning/OOM for a composite multimodal submodule versus 4-bit device_map/max_memory dispatching a model to CPU. They should not be merged as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 46822 is the better canonical anchor: it is open, specific, and describes a concrete zero.Init partitioning failure; issue 46823 is a separate CPU-dispatch regression.", + "canonical_pr_reason": null, + "best_issue_reason": "46822 is the strongest standalone issue report because it clearly identifies the affected model path and failure mode (OOM during multi-GPU load).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46822", + "right": "issue:46823", + "accept": false, + "reason": "Different subsystems and failure modes: ZeRO-3 composite submodule partitioning/OOM vs 4-bit auto device mapping with max_memory causing CPU placement and ValueError. 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": 24, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46906-13", + "nodes": [ + "issue:46906", + "issue:46940", + "issue:46948", + "issue:46969", + "issue:46985", + "issue:46998", + "issue:47037", + "issue:47049", + "issue:47053", + "issue:47091", + "issue:47102", + "issue:47120", + "issue:47182" + ], + "soft_pairs": [ + "issue:46906|issue:46940", + "issue:46906|issue:46948", + "issue:46906|issue:46969", + "issue:46906|issue:46985", + "issue:46906|issue:47037", + "issue:46940|issue:46948", + "issue:46940|issue:46969", + "issue:46940|issue:46985", + "issue:46940|issue:47049", + "issue:46948|issue:46969", + "issue:46948|issue:46985", + "issue:46948|issue:46998", + "issue:46969|issue:46985", + "issue:46969|issue:47037", + "issue:46969|issue:47091", + "issue:46985|issue:47037", + "issue:46985|issue:47049", + "issue:46998|issue:47053", + "issue:46998|issue:47037", + "issue:46998|issue:47049", + "issue:46985|issue:46998", + "issue:46998|issue:47091", + "issue:47037|issue:47049", + "issue:47037|issue:47091", + "issue:47037|issue:47053", + "issue:47049|issue:47053", + "issue:47049|issue:47091", + "issue:47053|issue:47091", + "issue:47053|issue:47102", + "issue:47091|issue:47102", + "issue:46985|issue:47091", + "issue:47037|issue:47102", + "issue:46985|issue:47102", + "issue:47049|issue:47102", + "issue:47120|issue:47182", + "issue:47102|issue:47120", + "issue:47053|issue:47120", + "issue:47049|issue:47120", + "issue:46998|issue:47120", + "issue:47102|issue:47182", + "issue:47053|issue:47182", + "issue:47037|issue:47182", + "issue:47091|issue:47182" + ], + "prepared_review_unit_hash": "9372e65054c9a0d9b394042362aa8daf8eccfc1f91d0159d7f76515c68c1fcf9", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 22741, + "estimated_input_tokens": 5686, + "estimated_eval_tokens": 11628 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 22741, + "estimated_input_tokens": 5686, + "estimated_eval_tokens": 11628 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are separate daily \"serge\" integration-failure triage issues, not duplicates of one underlying bug. The high similarity comes from a shared template and date-stamped titles, not evidence of the same defect.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:47182 is the latest open triage thread and the best representative of the current series, but it should not be treated as a duplicate of the earlier dated reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46906", + "right": "issue:46940", + "accept": false, + "reason": "Different dated triage reports; similarity is template-based, not the same underlying failure." + }, + { + "left": "issue:46906", + "right": "issue:46948", + "accept": false, + "reason": "Different dated triage reports; no evidence they track one shared bug." + }, + { + "left": "issue:46906", + "right": "issue:46969", + "accept": false, + "reason": "Separate daily reports for different dates, not duplicates." + }, + { + "left": "issue:46906", + "right": "issue:46985", + "accept": false, + "reason": "Same triage template, but each issue is a distinct day\u2019s report." + }, + { + "left": "issue:46906", + "right": "issue:47037", + "accept": false, + "reason": "The only overlap is the recurring triage wording; these are distinct reports." + }, + { + "left": "issue:46940", + "right": "issue:46948", + "accept": false, + "reason": "Different daily triage issues; no shared concrete defect is shown." + }, + { + "left": "issue:46940", + "right": "issue:46969", + "accept": false, + "reason": "Each issue corresponds to a different date's integration triage." + }, + { + "left": "issue:46940", + "right": "issue:46985", + "accept": false, + "reason": "Templated similarity only; not enough to call the same bug." + }, + { + "left": "issue:46940", + "right": "issue:47049", + "accept": false, + "reason": "Separate dated triage threads, not one mergeable duplicate." + }, + { + "left": "issue:46948", + "right": "issue:46969", + "accept": false, + "reason": "Different day triage reports with no concrete shared failure details." + }, + { + "left": "issue:46948", + "right": "issue:46985", + "accept": false, + "reason": "These appear to be independent daily reports, not duplicates." + }, + { + "left": "issue:46948", + "right": "issue:46998", + "accept": false, + "reason": "No evidence of the same underlying issue beyond the template." + }, + { + "left": "issue:46969", + "right": "issue:46985", + "accept": false, + "reason": "Adjacent daily triage issues, but still separate reports." + }, + { + "left": "issue:46969", + "right": "issue:47037", + "accept": false, + "reason": "The date-stamped triage pattern indicates distinct occurrences." + }, + { + "left": "issue:46969", + "right": "issue:47091", + "accept": false, + "reason": "Different daily integration-failure triage threads." + }, + { + "left": "issue:46985", + "right": "issue:47037", + "accept": false, + "reason": "Same subsystem and template, but not the same concrete problem." + }, + { + "left": "issue:46985", + "right": "issue:47049", + "accept": false, + "reason": "Separate dated triage reports; similarity is superficial." + }, + { + "left": "issue:46998", + "right": "issue:47053", + "accept": false, + "reason": "Both are triage issues, but for different dates and likely different failures." + }, + { + "left": "issue:46998", + "right": "issue:47037", + "accept": false, + "reason": "No sign these are the same bug; just recurring triage notes." + }, + { + "left": "issue:46998", + "right": "issue:47049", + "accept": false, + "reason": "Different daily report entries, not a duplicate pair." + }, + { + "left": "issue:46985", + "right": "issue:46998", + "accept": false, + "reason": "Date-stamped triage issues; no shared underlying fix target." + }, + { + "left": "issue:46998", + "right": "issue:47091", + "accept": false, + "reason": "Same template, but distinct integration-failure reports." + }, + { + "left": "issue:47037", + "right": "issue:47049", + "accept": false, + "reason": "Separate daily triage issues rather than one defect." + }, + { + "left": "issue:47037", + "right": "issue:47091", + "accept": false, + "reason": "No evidence of one shared code-path problem." + }, + { + "left": "issue:47037", + "right": "issue:47053", + "accept": false, + "reason": "These are recurring triage threads, not duplicates." + }, + { + "left": "issue:47049", + "right": "issue:47053", + "accept": false, + "reason": "Different dates and no concrete bug linkage." + }, + { + "left": "issue:47049", + "right": "issue:47091", + "accept": false, + "reason": "Template similarity only; not the same issue." + }, + { + "left": "issue:47053", + "right": "issue:47091", + "accept": false, + "reason": "Separate daily triage reports with no shared defect details." + }, + { + "left": "issue:47053", + "right": "issue:47102", + "accept": false, + "reason": "Not enough evidence that they describe the same failure." + }, + { + "left": "issue:47091", + "right": "issue:47102", + "accept": false, + "reason": "Same recurring triage wording, but different reports." + }, + { + "left": "issue:46985", + "right": "issue:47091", + "accept": false, + "reason": "Different day triage threads; no concrete duplicate signal." + }, + { + "left": "issue:47037", + "right": "issue:47102", + "accept": false, + "reason": "Separate daily failure triage issues." + }, + { + "left": "issue:46985", + "right": "issue:47102", + "accept": false, + "reason": "These look like successive reports, not one bug." + }, + { + "left": "issue:47049", + "right": "issue:47102", + "accept": false, + "reason": "Distinct dated triage issues with only template overlap." + }, + { + "left": "issue:47120", + "right": "issue:47182", + "accept": false, + "reason": "Both are triage reports for different dates; not the same underlying change/bug." + }, + { + "left": "issue:47102", + "right": "issue:47120", + "accept": false, + "reason": "Different daily integration-failure triage tickets." + }, + { + "left": "issue:47053", + "right": "issue:47120", + "accept": false, + "reason": "No evidence these are duplicate defect reports." + }, + { + "left": "issue:47049", + "right": "issue:47120", + "accept": false, + "reason": "Same series, but each issue tracks a separate day." + }, + { + "left": "issue:46998", + "right": "issue:47120", + "accept": false, + "reason": "Templated similarity only; not a mergeable duplicate." + }, + { + "left": "issue:47102", + "right": "issue:47182", + "accept": false, + "reason": "Latest two reports are still separate daily triage issues." + }, + { + "left": "issue:47053", + "right": "issue:47182", + "accept": false, + "reason": "Different dates and no concrete shared failure described." + }, + { + "left": "issue:47037", + "right": "issue:47182", + "accept": false, + "reason": "These are individual daily triage threads, not duplicates." + }, + { + "left": "issue:47091", + "right": "issue:47182", + "accept": false, + "reason": "No indication of the same underlying bug; just recurring triage." + } + ] + }, + "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": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46983-2", + "nodes": [ + "issue:46983", + "issue:46984" + ], + "soft_pairs": [ + "issue:46983|issue:46984" + ], + "prepared_review_unit_hash": "f93d402c5ca82535b57513beb931925017ee2a8e27d914b7738cdd00eb6e2f85", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1492, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1492, + "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: identical title, very similar bodies, and high text similarity indicate the same RWKV7 model support request.", + "confidence": 0.98, + "canonical_issue_reason": "Issue #46984 is the better canonical because it has slightly more discussion activity, suggesting it is the more active/complete thread for the same request.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #46984 is the best representative issue: same underlying request as #46983, with a bit more discussion and no evidence of a separate scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46983", + "right": "issue:46984", + "accept": true, + "reason": "Same title and highly similar issue bodies; both are clearly the same RWKV7 model support request." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The conclusion is grounded in the packet: both issues share the same title, have high textual similarity, and issue #46984 has slightly more discussion activity. The representative/canonical choice is conservative and supported by the available evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 26, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-47007-2", + "nodes": [ + "issue:47007", + "issue:47015" + ], + "soft_pairs": [ + "issue:47007|issue:47015" + ], + "prepared_review_unit_hash": "596babec546b9ffe3bdab8ad44908df6089afedd17bd3f3d0e26f2d2264e4f08", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1626, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1626, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both closed issues are broadly about AI/LLM-driven vulnerability discovery, but they read like different topic/proposal posts rather than the same underlying bug or change. With no shared targets, filenames, or concrete code-path evidence, I would not merge them as duplicates.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 47015 is the more direct, self-contained statement of the cybersecurity-vulnerability-discovery topic, so it is the better anchor if this cluster must be represented by a single issue.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47015 is the clearest and most specific of the two titles, making it the better representative for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47007", + "right": "issue:47015", + "accept": false, + "reason": "Only soft textual similarity is present. The titles describe different framings of AI-assisted vulnerability discovery, not the same concrete bug, code path, or change, and there is no supporting issue/filename evidence." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 27, + "component_count": 27, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-47138-2", + "nodes": [ + "issue:47138", + "issue:47139" + ], + "soft_pairs": [ + "issue:47138|issue:47139" + ], + "prepared_review_unit_hash": "4b4722fe03cc24e217115fd910af081d25be22a1b84076886312a98de6dba4cd", + "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": "47139 is the better canonical issue because it shares the exact title with 47138 and has more inbound references, but the duplicate link itself is only weakly supported by the available metadata.", + "confidence": 0.89, + "canonical_issue_reason": "47139 has the same title as 47138 and already has more inbound references, so it is the better canonical target for consolidation.", + "canonical_pr_reason": null, + "best_issue_reason": "47139 is the best issue candidate because it matches the same reported problem and is the more established reference point.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47138", + "right": "issue:47139", + "accept": false, + "reason": "The titles are identical, but there is not enough additional linkage or distinguishing detail in the packet to confidently treat them as the same underlying report." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The packet supports that the two issues have identical titles and that 47139 has more inbound references, so picking 47139 as the canonical target is reasonable. However, the analyst overstates the evidence by calling this a \"clear duplicate\" / \"evidently the same report\" based only on title match and very limited metadata. The soft-edge verdict should stay more conservative unless there is stronger linking evidence." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 33, + "review_unit_index": 1, + "review_unit_count": 3, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-54", + "nodes": [ + "pull_request:43451", + "pull_request:44517", + "pull_request:44775", + "pull_request:46196", + "pull_request:46266", + "pull_request:46267", + "pull_request:46376", + "pull_request:46395", + "pull_request:46834", + "pull_request:46836", + "pull_request:46895", + "pull_request:46903", + "pull_request:46912", + "pull_request:46954", + "pull_request:46957", + "pull_request:46958", + "pull_request:46974", + "pull_request:46995", + "pull_request:47003", + "pull_request:47034", + "pull_request:47035", + "pull_request:47069", + "pull_request:47072", + "pull_request:47077", + "pull_request:47078", + "pull_request:47082", + "pull_request:47089", + "pull_request:47090", + "pull_request:47093", + "pull_request:47108", + "pull_request:47113", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47173", + "pull_request:47175", + "pull_request:47178", + "pull_request:47179", + "pull_request:47181", + "pull_request:47184", + "pull_request:47202", + "pull_request:47204", + "pull_request:47207", + "pull_request:47210", + "pull_request:47212", + "pull_request:47216", + "pull_request:47218", + "pull_request:47223" + ], + "soft_pairs": [ + "pull_request:46834|pull_request:47223", + "pull_request:46903|pull_request:47090", + "pull_request:46903|pull_request:46958", + "pull_request:46954|pull_request:46958", + "pull_request:46958|pull_request:46974", + "pull_request:47072|pull_request:47089", + "pull_request:47202|pull_request:47204", + "pull_request:47077|pull_request:47078", + "pull_request:47108|pull_request:47202", + "pull_request:47077|pull_request:47212", + "pull_request:47173|pull_request:47184", + "pull_request:46266|pull_request:46957", + "pull_request:47166|pull_request:47173", + "pull_request:44517|pull_request:47082", + "pull_request:47035|pull_request:47202", + "pull_request:46912|pull_request:47179", + "pull_request:47034|pull_request:47035", + "pull_request:47163|pull_request:47173", + "pull_request:46995|pull_request:47163", + "pull_request:46196|pull_request:47210", + "pull_request:46376|pull_request:47202", + "pull_request:47003|pull_request:47204", + "pull_request:46395|pull_request:47181", + "pull_request:47069|pull_request:47089", + "pull_request:46196|pull_request:46995", + "pull_request:47089|pull_request:47113", + "pull_request:46376|pull_request:46995", + "pull_request:46912|pull_request:47173", + "pull_request:46196|pull_request:47202", + "pull_request:46995|pull_request:47181", + "pull_request:47207|pull_request:47218", + "pull_request:44775|pull_request:47077", + "pull_request:46995|pull_request:47143", + "pull_request:47082|pull_request:47108", + "pull_request:46196|pull_request:47179", + "pull_request:46267|pull_request:47181", + "pull_request:47093|pull_request:47181", + "pull_request:47179|pull_request:47210", + "pull_request:44517|pull_request:46376", + "pull_request:47204|pull_request:47207", + "pull_request:44775|pull_request:47181", + "pull_request:47077|pull_request:47210", + "pull_request:47093|pull_request:47143", + "pull_request:47175|pull_request:47179", + "pull_request:47093|pull_request:47163", + "pull_request:47163|pull_request:47179", + "pull_request:46266|pull_request:47216", + "pull_request:46836|pull_request:47218", + "pull_request:46376|pull_request:47218", + "pull_request:44775|pull_request:46836", + "pull_request:47108|pull_request:47216", + "pull_request:47082|pull_request:47210", + "pull_request:46912|pull_request:47078", + "pull_request:46836|pull_request:47178", + "pull_request:47108|pull_request:47204", + "pull_request:46957|pull_request:47216", + "pull_request:47082|pull_request:47181", + "pull_request:47166|pull_request:47218", + "pull_request:46912|pull_request:47082", + "pull_request:47035|pull_request:47089", + "pull_request:46895|pull_request:47207", + "pull_request:44775|pull_request:46376", + "pull_request:43451|pull_request:44775" + ], + "prepared_review_unit_hash": "e979a438321fe272482268239b95d0b9d061dc4068d12a98fc1b5514af4e3c9c", + "component_budget": { + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 134113, + "estimated_input_tokens": 33529, + "estimated_eval_tokens": 67314 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a heterogeneous cluster: most PRs are unrelated model additions, docs, test updates, or infrastructure changes. The only clear duplicate subgroup is the CohereASR training-loss double-shift fixes, which all target the same loss path; most other soft edges share a subsystem but not the same concrete change.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47090 is the cleanest representative of the duplicate CohereASR loss bug: it fixes the exact training-loss double-shift in the target model, keeps the loss helper semantics intact, and is more focused than the broader or rougher variants.", + "best_issue_reason": null, + "best_pr_reason": "PR 47090 is the strongest global PR candidate because it is narrowly scoped, well explained, and fixes the actual CohereASR code path; the nearby alternatives are either broader multi-model changes or less precise implementations.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46834", + "right": "pull_request:47223", + "accept": false, + "reason": "Same model area and issue target, but these fix different bugs: a method-name typo in forward vs config attribute placeholders/tests." + }, + { + "left": "pull_request:46903", + "right": "pull_request:47090", + "accept": true, + "reason": "Both fix the same CohereASR training-loss double-shift bug in the same code path; the implementation details differ, but the underlying change is the same." + }, + { + "left": "pull_request:46903", + "right": "pull_request:46958", + "accept": true, + "reason": "Same CohereASR loss double-shift bug and same affected forward path; these are alternative implementations of the same fix." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46958", + "accept": false, + "reason": "Unrelated changes in different models: Pytest-v9 fp32-module handling vs CohereASR loss shifting." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46974", + "accept": false, + "reason": "Different models and problems: CohereASR training loss vs a MusicGen generate typo." + }, + { + "left": "pull_request:47072", + "right": "pull_request:47089", + "accept": false, + "reason": "Different utilities and bugs: chat-template enum parsing vs AqlmConfig type-error messages." + }, + { + "left": "pull_request:47202", + "right": "pull_request:47204", + "accept": false, + "reason": "Both touch docker/CI, but one removes the old image builder while the other adds exporter dependencies to torch-light; not the same change." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47078", + "accept": false, + "reason": "Both are Japanese docs edits, but one is a DistilBERT translation PR and the other is wording cleanup; not the same underlying change." + }, + { + "left": "pull_request:47108", + "right": "pull_request:47202", + "accept": false, + "reason": "Doc-builder comment tweak vs docker image builder deletion; unrelated changes." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47212", + "accept": false, + "reason": "Japanese docs translation vs kv-cache documentation updates; different content and purpose." + }, + { + "left": "pull_request:47173", + "right": "pull_request:47184", + "accept": false, + "reason": "Both are test updates for accelerator-related behavior, but they cover different models and different test expectations." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46957", + "accept": false, + "reason": "Different code paths: DINO conversion/mapping updates vs generation-time FP8 kernel routing." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47173", + "accept": false, + "reason": "Unrelated test updates in different areas: quantization expectations vs XPU expectation fixes." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47082", + "accept": false, + "reason": "Different features entirely: Qwen3-TTS model/docs vs RoMa model support." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47202", + "accept": false, + "reason": "Torch-compile Neuron defaulting and docker image cleanup are unrelated changes." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47179", + "accept": false, + "reason": "RoPE test improvements are unrelated to Granite-SWA model support." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47035", + "accept": false, + "reason": "Related Neuron work, but one adds a training example while the other changes the default compile backend; not the same change." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47173", + "accept": false, + "reason": "Both are accelerator/XPU-oriented tests, but they are for different models and different behaviors." + }, + { + "left": "pull_request:46995", + "right": "pull_request:47163", + "accept": false, + "reason": "Flash-linear-attention availability change vs paged-attention XPU tests; different bugs." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47210", + "accept": false, + "reason": "Both are attention-related, but one is broad SDPA dispatch support and the other is packed-row varlen attention routing." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47202", + "accept": false, + "reason": "Continuous-batching multimodal support is unrelated to docker image builder deletion." + }, + { + "left": "pull_request:46395", + "right": "pull_request:47181", + "accept": false, + "reason": "Different model additions: Param2MoE vs Cosmos3 Edge." + }, + { + "left": "pull_request:47069", + "right": "pull_request:47089", + "accept": false, + "reason": "Num-labels None typing fix is unrelated to AqlmConfig error-message text." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "Different attention features and code paths; not the same bug." + }, + { + "left": "pull_request:47089", + "right": "pull_request:47113", + "accept": false, + "reason": "AqlmConfig validation wording is unrelated to audio feature/proc requirement decorators." + }, + { + "left": "pull_request:46376", + "right": "pull_request:46995", + "accept": false, + "reason": "Continuous batching and flash-linear-attention availability are separate changes." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47173", + "accept": false, + "reason": "RoPE test coverage vs XPU test expectation updates are unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47202", + "accept": false, + "reason": "SDPA dispatch support and docker image cleanup do not share the same underlying change." + }, + { + "left": "pull_request:46995", + "right": "pull_request:47181", + "accept": false, + "reason": "Flash-linear-attention availability and Cosmos3 Edge model support are unrelated." + }, + { + "left": "pull_request:47207", + "right": "pull_request:47218", + "accept": false, + "reason": "Image padding logic fix vs test image-loading refactor; not the same bug." + }, + { + "left": "pull_request:44775", + "right": "pull_request:47077", + "accept": false, + "reason": "N-D parallelism docs and Japanese DistilBERT docs are unrelated documentation updates." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47216", + "accept": false, + "reason": "FP8 kernel optimization vs compressed-tensors fp8 support; both quantization-adjacent, but different concrete changes." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47181", + "accept": false, + "reason": "RoMa keypoint matcher support and Cosmos3 Edge support are separate model additions." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47179", + "accept": false, + "reason": "Olmo3 CI test expectation updates are unrelated to Granite-SWA model support." + }, + { + "left": "pull_request:47093", + "right": "pull_request:47143", + "accept": false, + "reason": "OpenPangu-V2 and NeuCodec are different model-addition PRs." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47179", + "accept": false, + "reason": "Paged-attention XPU tests and Granite-SWA support are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47216", + "accept": false, + "reason": "Modernizing Dino conversion/mapping and compressed-tensors fp8 support are different changes." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47218", + "accept": false, + "reason": "Unlimited OCR model support and image-loading test refactor are unrelated." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47218", + "accept": false, + "reason": "Continuous batching multimodal support and test image loading refactor are unrelated." + }, + { + "left": "pull_request:44775", + "right": "pull_request:46836", + "accept": false, + "reason": "Docs for parallelism and Unlimited OCR model support are unrelated." + }, + { + "left": "pull_request:47108", + "right": "pull_request:47216", + "accept": false, + "reason": "Doc-builder pin comment cleanup is unrelated to compressed-tensors fp8 work." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47216", + "accept": false, + "reason": "Same broad quantization area, but different concrete changes and code paths." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47181", + "accept": false, + "reason": "Different model additions with no shared bug/change." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47218", + "accept": false, + "reason": "Quantization test expectation updates are unrelated to image-loading refactors." + }, + { + "left": "pull_request:43451", + "right": "pull_request:44775", + "accept": false, + "reason": "Molmo2 model addition and N-D parallelism docs are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 33, + "review_unit_index": 2, + "review_unit_count": 3, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-54", + "nodes": [ + "pull_request:40546", + "pull_request:43451", + "pull_request:44517", + "pull_request:46196", + "pull_request:46266", + "pull_request:46376", + "pull_request:46509", + "pull_request:46658", + "pull_request:46836", + "pull_request:46895", + "pull_request:46912", + "pull_request:46957", + "pull_request:46958", + "pull_request:47003", + "pull_request:47034", + "pull_request:47035", + "pull_request:47069", + "pull_request:47077", + "pull_request:47078", + "pull_request:47082", + "pull_request:47089", + "pull_request:47090", + "pull_request:47096", + "pull_request:47108", + "pull_request:47113", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47173", + "pull_request:47175", + "pull_request:47178", + "pull_request:47181", + "pull_request:47184", + "pull_request:47198", + "pull_request:47204", + "pull_request:47207", + "pull_request:47210", + "pull_request:47216", + "pull_request:47218", + "pull_request:47219", + "pull_request:47223" + ], + "soft_pairs": [ + "pull_request:47035|pull_request:47198", + "pull_request:47035|pull_request:47204", + "pull_request:46376|pull_request:47173", + "pull_request:46196|pull_request:46376", + "pull_request:47113|pull_request:47204", + "pull_request:46196|pull_request:47223", + "pull_request:47089|pull_request:47223", + "pull_request:46266|pull_request:46376", + "pull_request:47035|pull_request:47113", + "pull_request:47035|pull_request:47069", + "pull_request:47069|pull_request:47113", + "pull_request:46266|pull_request:47034", + "pull_request:47034|pull_request:47108", + "pull_request:47069|pull_request:47082", + "pull_request:46376|pull_request:47207", + "pull_request:47003|pull_request:47178", + "pull_request:43451|pull_request:47078", + "pull_request:47078|pull_request:47096", + "pull_request:46912|pull_request:47113", + "pull_request:47003|pull_request:47078", + "pull_request:47113|pull_request:47175", + "pull_request:47175|pull_request:47207", + "pull_request:44517|pull_request:47181", + "pull_request:47113|pull_request:47223", + "pull_request:43451|pull_request:46266", + "pull_request:46376|pull_request:47198", + "pull_request:43451|pull_request:44517", + "pull_request:47077|pull_request:47216", + "pull_request:47069|pull_request:47175", + "pull_request:44517|pull_request:47143", + "pull_request:46895|pull_request:47223", + "pull_request:46658|pull_request:46895", + "pull_request:43451|pull_request:46957", + "pull_request:46957|pull_request:47096", + "pull_request:46376|pull_request:47184", + "pull_request:47077|pull_request:47198", + "pull_request:46957|pull_request:47003", + "pull_request:47175|pull_request:47223", + "pull_request:46895|pull_request:47090", + "pull_request:46958|pull_request:47090", + "pull_request:46895|pull_request:46958", + "pull_request:47218|pull_request:47219", + "pull_request:47166|pull_request:47184", + "pull_request:46912|pull_request:47175", + "pull_request:46658|pull_request:46836", + "pull_request:47163|pull_request:47166", + "pull_request:47163|pull_request:47184", + "pull_request:46509|pull_request:47143", + "pull_request:46509|pull_request:47181", + "pull_request:46509|pull_request:47163", + "pull_request:46376|pull_request:46509", + "pull_request:43451|pull_request:46836", + "pull_request:46376|pull_request:47181", + "pull_request:46509|pull_request:46836", + "pull_request:46836|pull_request:47143", + "pull_request:46836|pull_request:47163", + "pull_request:47143|pull_request:47181", + "pull_request:46376|pull_request:47143", + "pull_request:43451|pull_request:46509", + "pull_request:47163|pull_request:47181", + "pull_request:46376|pull_request:47163", + "pull_request:46376|pull_request:47034", + "pull_request:46658|pull_request:47143", + "pull_request:47163|pull_request:47210", + "pull_request:47143|pull_request:47163", + "pull_request:47077|pull_request:47181", + "pull_request:47034|pull_request:47216", + "pull_request:47143|pull_request:47210", + "pull_request:47181|pull_request:47198", + "pull_request:40546|pull_request:43451", + "pull_request:43451|pull_request:47175", + "pull_request:47096|pull_request:47175" + ], + "prepared_review_unit_hash": "d8b229da8b05a2988ceda914eba9faf4d7182ad4062c819c2d616530fab3c55c", + "component_budget": { + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 + }, + "budget": { + "node_count": 41, + "item_count": 41, + "soft_pair_count": 72, + "serialized_chars": 127033, + "estimated_input_tokens": 31759, + "estimated_eval_tokens": 63774 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated PRs (new model integrations, docs tweaks, test updates, exporter changes). Only two true duplicate mini-clusters stand out: the CohereASR training-loss double-shift fixes and the Aimv2 image-loading test migrations. Everything else is only broadly similar at the subsystem level, not the same concrete change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single PR cleanly represents the whole mixed cluster; among the real duplicate subgroup, PR 47090 is the clearest canonical representative because it is the most explicit and complete CohereASR double-shift fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 47090 is the best global PR fit among the actual duplicates: it fixes a concrete, well-scoped code-path bug with an explicit issue tie-in, while most other items are unrelated or only superficially similar.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47035", + "right": "pull_request:47198", + "accept": false, + "reason": "Both touch backend/model plumbing, but they fix different features and code paths." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47204", + "accept": false, + "reason": "Neuron compile default vs exporter dependencies are unrelated changes." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47173", + "accept": false, + "reason": "Both are test/infrastructure oriented, but they address different subsystems and failures." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46376", + "accept": false, + "reason": "Both are generation-related, but one is SDPA dispatch and the other is continuous batching support." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47204", + "accept": false, + "reason": "Audio requirement fixes vs docker exporter dependencies are not the same change." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47223", + "accept": false, + "reason": "Unrelated generation/backend work; no shared concrete bug or change." + }, + { + "left": "pull_request:47089", + "right": "pull_request:47223", + "accept": false, + "reason": "Quantization config type-check messages vs TimesFM AttributeError fix are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46376", + "accept": false, + "reason": "Different modernization tasks in different areas; only broad similarity." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47113", + "accept": false, + "reason": "Neuron compile default and audio feature requirements are unrelated." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47069", + "accept": false, + "reason": "Trainium compile backend vs num_labels typing fix are not the same bug/change." + }, + { + "left": "pull_request:47069", + "right": "pull_request:47113", + "accept": false, + "reason": "Configuration typing and audio requirements are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47034", + "accept": false, + "reason": "Both are Trainium/backend adjacent, but one is a model-support PR and the other is an experiment script." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47108", + "accept": false, + "reason": "Trainium experiment code and doc-builder pin comments do not share the same underlying change." + }, + { + "left": "pull_request:47069", + "right": "pull_request:47082", + "accept": false, + "reason": "A config property typing fix and a new keypoint matcher are unrelated." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47207", + "accept": false, + "reason": "Continuous batching multimodal support and Swin2SR padding fix are unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47178", + "accept": false, + "reason": "OpenVINO exporter support and SigLIP path fixes are different code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47078", + "accept": false, + "reason": "Both are docs-heavy, but one adds a model and the other tweaks Japanese wording." + }, + { + "left": "pull_request:47078", + "right": "pull_request:47096", + "accept": false, + "reason": "Japanese docs wording/model text changes are unrelated to Kimi exportability." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47113", + "accept": false, + "reason": "RoPE test updates and audio feature/proc requirement fixes are unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47078", + "accept": false, + "reason": "Exporter support and Japanese doc wording updates are not the same change." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47175", + "accept": false, + "reason": "Audio requirements fixes and Olmo3 CI expectation updates are unrelated." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47207", + "accept": false, + "reason": "Different test/model areas with no shared bug or patch." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47181", + "accept": false, + "reason": "Both add model docs/support, but for different models and different code paths." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47223", + "accept": false, + "reason": "Audio requirements and TimesFM window-size fix are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46266", + "accept": false, + "reason": "Both are model-integration PRs, but they add different models and code." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47198", + "accept": false, + "reason": "Continuous batching and backend compatibility fixes are different changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:44517", + "accept": false, + "reason": "Different model additions; only the broad pattern is shared." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47216", + "accept": false, + "reason": "Japanese docs translation work and compressed-tensors quantization docs/features are unrelated." + }, + { + "left": "pull_request:47069", + "right": "pull_request:47175", + "accept": false, + "reason": "A config property typing fix and CI expectation refresh are unrelated." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47143", + "accept": false, + "reason": "Different model-support additions with no shared concrete implementation." + }, + { + "left": "pull_request:46895", + "right": "pull_request:47090", + "accept": true, + "reason": "Same CohereASR training-loss double-shift bug; same files and same fix direction." + }, + { + "left": "pull_request:46958", + "right": "pull_request:47090", + "accept": true, + "reason": "Same CohereASR double-shift loss bug; 47090 is the cleaner follow-up fix for the same code path." + }, + { + "left": "pull_request:46895", + "right": "pull_request:46958", + "accept": true, + "reason": "Both target the same CohereASR loss double-shift issue in the same files." + }, + { + "left": "pull_request:47218", + "right": "pull_request:47219", + "accept": true, + "reason": "Both migrate the same image-loading tests to HF Hub fixtures/helper loading and appear to be the same underlying test change." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47184", + "accept": false, + "reason": "Both are test updates, but one is quantization coverage and the other is accelerator/device-map test adjustments." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47175", + "accept": false, + "reason": "Different fixes: RoPE test coverage vs Olmo3 CI expectation updates." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46836", + "accept": false, + "reason": "Step3.7 flash work and Unlimited OCR are unrelated model additions." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47166", + "accept": false, + "reason": "Paged-attention XPU test support and quantization test updates are not the same underlying change." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47184", + "accept": false, + "reason": "Different test suites and different hardware-targeted adjustments." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47143", + "accept": false, + "reason": "Nemotron Omni support and NeuCodec support are different model integrations." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47181", + "accept": false, + "reason": "Different model families and different integration code." + }, + { + "left": "pull_request:46376", + "right": "pull_request:46509", + "accept": false, + "reason": "Continuous batching changes and Nemotron Omni model support are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46836", + "accept": false, + "reason": "Different model additions with different subsystems and files." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47181", + "accept": false, + "reason": "Continuous batching and Cosmos3 Edge model support are unrelated." + }, + { + "left": "pull_request:46509", + "right": "pull_request:46836", + "accept": false, + "reason": "Nemotron Omni and Unlimited OCR are unrelated model additions." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47143", + "accept": false, + "reason": "Unlimited OCR support and NeuCodec support are different model integrations." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47181", + "accept": false, + "reason": "Different model additions; same broad pattern, not the same change." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47163", + "accept": false, + "reason": "Continuous batching internals vs paged-attention test coverage are different changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46509", + "accept": false, + "reason": "Different model-support PRs for different architectures." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47181", + "accept": false, + "reason": "Paged-attention tests and Cosmos3 Edge model support are unrelated." + }, + { + "left": "pull_request:46376", + "right": "pull_request:47034", + "accept": false, + "reason": "Continuous batching work and Trainium SFT experiment code are unrelated." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47143", + "accept": false, + "reason": "Step3.7 flash tooling and NeuCodec model support are unrelated." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47210", + "accept": false, + "reason": "Paged-attention XPU tests and packed SDPA varlen routing are different attention changes." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47210", + "accept": false, + "reason": "NeuCodec support and packed SDPA routing are unrelated." + }, + { + "left": "pull_request:47181", + "right": "pull_request:47198", + "accept": false, + "reason": "Different model-support work; no shared concrete bug or code path." + }, + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "Both are large model-support PRs, but they add different models and unrelated code." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47175", + "accept": false, + "reason": "Molmo2 model support and Olmo3 CI expectations are unrelated." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47175", + "accept": false, + "reason": "Exportability fixes for Kimi and Olmo3 CI updates 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": 1, + "component_count": 33, + "review_unit_index": 3, + "review_unit_count": 3, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-54", + "nodes": [ + "pull_request:43451", + "pull_request:46509", + "pull_request:46658", + "pull_request:47003", + "pull_request:47034", + "pull_request:47096", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47175", + "pull_request:47181", + "pull_request:47184", + "pull_request:47198", + "pull_request:47223" + ], + "soft_pairs": [ + "pull_request:47003|pull_request:47175", + "pull_request:43451|pull_request:46658", + "pull_request:46658|pull_request:47096", + "pull_request:46658|pull_request:47003", + "pull_request:43451|pull_request:47143", + "pull_request:43451|pull_request:47163", + "pull_request:43451|pull_request:47198", + "pull_request:43451|pull_request:47096", + "pull_request:43451|pull_request:47003", + "pull_request:47003|pull_request:47096", + "pull_request:43451|pull_request:47034", + "pull_request:47034|pull_request:47096", + "pull_request:43451|pull_request:47166", + "pull_request:47096|pull_request:47166", + "pull_request:43451|pull_request:47184", + "pull_request:43451|pull_request:47223", + "pull_request:47096|pull_request:47223", + "pull_request:43451|pull_request:47181", + "pull_request:47003|pull_request:47223", + "pull_request:47096|pull_request:47143", + "pull_request:47003|pull_request:47143", + "pull_request:47096|pull_request:47163", + "pull_request:47003|pull_request:47163", + "pull_request:46509|pull_request:47096" + ], + "prepared_review_unit_hash": "18548f9e86c9b6139b14daaf6430cfd514b2be221aee1eb91da233410030dd9a", + "component_budget": { + "node_count": 54, + "item_count": 54, + "soft_pair_count": 159, + "serialized_chars": 197087, + "estimated_input_tokens": 49272, + "estimated_eval_tokens": 98800 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 24, + "serialized_chars": 41303, + "estimated_input_tokens": 10326, + "estimated_eval_tokens": 20908 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are mostly unrelated: model additions, exporter work, test updates, and one backend/conversion fix. The shared similarity is broad Transformers plumbing, not a single underlying bug or change, so no soft edges should be merged.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true canonical PR emerges from this cluster; the items span unrelated features and tests. If one must be named, 43451 is the most substantial standalone model-add PR, but it is not a duplicate of the others.", + "best_issue_reason": null, + "best_pr_reason": "43451 is the strongest representative of the cluster only in the sense that it is the largest, most complete feature PR here; however, it should not be used as a duplicate sink for the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47003", + "right": "pull_request:47175", + "accept": false, + "reason": "OpenVINO exporter work is unrelated to Olmo3 CI test expectation updates; different code paths and different goals." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46658", + "accept": false, + "reason": "Molmo2 model support and Step-3.7 flash checkpoint conversion are separate model integrations with different files, code paths, and release targets." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47096", + "accept": false, + "reason": "Step-3.7 flash model conversion/loading is unrelated to Kimi exportability changes; they touch different models and mechanics." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47003", + "accept": false, + "reason": "Step-3.7 flash checkpoint tooling is not the same change as adding an OpenVINO exporter." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47143", + "accept": false, + "reason": "Molmo2 support and NeuCodec support are distinct model additions with different configs, processors, and tests." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47163", + "accept": false, + "reason": "Molmo2 model support is unrelated to adding XPU paged-attention tests for generation." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47198", + "accept": false, + "reason": "Molmo2 model addition does not match the GPTBigCode/GPTNeoX backend/export compatibility fix." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47096", + "accept": false, + "reason": "Molmo2 model support is unrelated to Kimi exportability and attention-backend changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47003", + "accept": false, + "reason": "Model addition work for Molmo2 is a different change than introducing OpenVINO exporter support." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47096", + "accept": false, + "reason": "OpenVINO exporter support and Kimi exportability are both exporter-related only at a very broad level; they fix different code paths and should not be merged as duplicates." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47034", + "accept": false, + "reason": "Molmo2 model support is unrelated to the Neuron SFT experiment and training-script changes." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47096", + "accept": false, + "reason": "Neuron finetuning experiment changes do not overlap with Kimi export/backend fixes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47166", + "accept": false, + "reason": "Molmo2 model support and quantization test expansions are different kinds of changes with no shared underlying bug." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47166", + "accept": false, + "reason": "Kimi export/backend work is not the same as GPTQ/MXFP4 test adjustments." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47184", + "accept": false, + "reason": "Molmo2 model support is unrelated to the Qwen3 Next multi-accelerator test annotation change." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47223", + "accept": false, + "reason": "Molmo2 model addition is unrelated to the TimesFM2.5 window_size AttributeError fix." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47223", + "accept": false, + "reason": "Kimi export/back-end changes and TimesFM2.5 config/test fixes address different models and failure modes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47181", + "accept": false, + "reason": "Molmo2 and Cosmos3 Edge are separate model-support PRs with different architectures and files." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47223", + "accept": false, + "reason": "OpenVINO exporter support is unrelated to the TimesFM2.5 AttributeError fix." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47143", + "accept": false, + "reason": "Kimi export/backend support and NeuCodec model support are unrelated changes." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47143", + "accept": false, + "reason": "OpenVINO exporter support does not describe the same underlying feature as NeuCodec model integration." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47163", + "accept": false, + "reason": "Kimi export/backend fixes are unrelated to paged-attention test coverage for XPU." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47163", + "accept": false, + "reason": "OpenVINO exporter work is not the same as adding XPU paged-attention tests." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47096", + "accept": false, + "reason": "Nemotron Omni support is a model-addition PR; Kimi exportability is a separate backend compatibility 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": 2, + "component_count": 33, + "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": "ff9566c255cab4a813d1ae96af9aa79ffb1502218d596d8be8f0b2054e7f3d6c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4674, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4674, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are unrelated feature additions in different model families: TimesFM covariate forecasting vs. DINOv3 image-classification support. They should not be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "If one PR must represent the set, #41224 is the better canonical PR because it is a narrower, concrete feature addition with clearer model-specific scope and substantial review activity.", + "best_issue_reason": null, + "best_pr_reason": "#41224 is the strongest standalone PR in this cluster: it targets a specific code path, includes docs/tests, and has a well-defined feature boundary.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different models and different changes: TimesFM forecasting with covariates vs. DINOv3 image classification support. No shared files, targets, or underlying bug/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": 3, + "component_count": 33, + "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": 33, + "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-41855-12", + "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" + ], + "soft_pairs": [ + "pull_request:41855|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "afdec1af5aa3c2511a2c414d5f7f685b48549a0aa87501e37592fc81e3586cd3", + "component_budget": { + "node_count": 12, + "item_count": 12, + "soft_pair_count": 17, + "serialized_chars": 31599, + "estimated_input_tokens": 7900, + "estimated_eval_tokens": 16056 + }, + "budget": { + "node_count": 12, + "item_count": 12, + "soft_pair_count": 17, + "serialized_chars": 31599, + "estimated_input_tokens": 7900, + "estimated_eval_tokens": 16056 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The only likely duplicate pair is the two `fixed_cross_entropy` PRs; the rest are unrelated changes that merely touch adjacent transformer subsystems or share broad themes like model integration/testing.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43251 is the best canonical choice because it directly targets the shared `fixed_cross_entropy` bug, explicitly links the tracking issue, and contains the more complete fix for forwarding supported `cross_entropy` kwargs.", + "best_issue_reason": null, + "best_pr_reason": "PR 43251 is the strongest representative of the cluster\u2019s only true duplicate change-set: same file, same function, same issue, and a superset-style implementation of the same fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Different problems: Mistral tokenizer API gaps vs. Executorch dynamic-shape export testing." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Unrelated: tokenizer method additions for Mistral vs. adding a new MobileLLM model." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Both are processor/model loading work, but they fix different models and different code paths." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "`fixed_cross_entropy` kwargs forwarding is unrelated to adding MobileLLM model support." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Cross-entropy loss plumbing is unrelated to Mistral4 model/test changes." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Dynamic export tests and Mistral4 fixes do not address the same underlying bug." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Executorch export coverage vs. new model addition are different changes." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New model onboarding and `cls.config` resolution ordering are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "MobileLLM introduction and Falcon output-interface refactoring are different code paths." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "Model addition vs. repository skill/documentation changes; not the same change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "Different subsystems and bug types; no shared concrete fix." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon output-collection refactor is not the same as Mistral4 test/model fixes." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon model refactor vs. model-creation skill PR are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test/model fixes and `layrnorm` typo renaming are different bugs with different affected files." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Mistral4 fixes are unrelated to the model-creation skill/large scaffold PR." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "No common concrete bug: Mistral4 fixes vs. `cls.config` resolution order." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 fixes are unrelated to the TDT loss kernel addition." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "Adding a model and adding a model-creation skill are not the same underlying change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Same function, same file, same tracking issue, and both aim to forward supported `cross_entropy` kwargs; they could plausibly be merged into one fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted soft-edge pair, pull_request:43251 and pull_request:43254, is well-supported: same file, same function, same tracked issue, and the diffs are mergeable as one bugfix. The other rejected pairs are described at a high level without overstating evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 33, + "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": "ea38cc0f59f7a0f06fbe7daa85cb3a14c0a50ec80956fc5eff3beb475ade9b90", + "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": "Both pull requests are broad GitHub Actions maintenance bumps, but they target different underlying upgrades: one is a Node 24 compatibility sweep, the other is a latest-version bump focused on Docker/build actions. They overlap on some workflow files, but not on the same concrete change or fix.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the clearer representative: it has a specific compatibility goal (Node 24) and a coherent set of workflow-action updates across many files.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the strongest standalone choice because its intent is explicit and it covers the broader workflow-action compatibility migration, whereas 42976 is a separate version bump for different actions.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Reject: both are workflow dependency updates, but they do not fix the same concrete problem. 42975 is about Node 24 compatibility for core GitHub actions; 42976 upgrades docker/build-push-action and related actions to newer versions. Shared workflow files alone are not enough to treat them as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 33, + "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": "258393b9b1ae6f53dcd74e6d7b9f497c2273b498cd204e63896882c54f3cedce", + "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 items are independent model-addition PRs (ViT NEPA, Molmo, CTSM, Step3p7). Their overlap is limited to shared Hugging Face boilerplate like docs and auto-registration, not the same code-path fix or feature.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: these are distinct model integrations with different architectures, files, and test surfaces, so there is no duplicate target among them.", + "best_issue_reason": null, + "best_pr_reason": "Step3p7 is the strongest standalone representative in this set because it has the broadest integration surface and highest review activity, but it is still a separate model addition rather than a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "ViT NEPA and Molmo are different model integrations; shared registry/docs changes are boilerplate and not evidence of the same underlying change." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "ViT NEPA vs CTSM: different model families and code paths. The overlap is only generic auto/docs wiring." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM are separate new-model PRs; they do not appear to fix or add the same concrete implementation." + }, + { + "left": "pull_request:45490", + "right": "pull_request:46562", + "accept": false, + "reason": "CTSM and Step3p7 both touch model-registration infrastructure, but they introduce different models and unrelated implementations, 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": 7, + "component_count": 33, + "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": "df636795354635f36383c927b5ec64f4e2a6a4f0ff4628b871089c6d4641bf02", + "component_budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19601, + "estimated_input_tokens": 4901, + "estimated_eval_tokens": 10058 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19601, + "estimated_input_tokens": 4901, + "estimated_eval_tokens": 10058 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Only the two YOLOS PRs are close enough to treat as the same underlying fix; the rest are unrelated test-only or subsystem-specific changes despite similar titles/labels.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46922 is the best representative because it addresses the concrete YOLOS device-mismatch path directly and also includes the model split guard, making it the more complete version of the 46886/46922 fix pair.", + "best_issue_reason": null, + "best_pr_reason": "PR 46922 is the strongest standalone PR: it targets the same YOLOS issue as 46886, fixes the runtime device placement bug in forward, and includes the supporting _no_split_modules change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Both are generic test-fix PRs, but they touch different models and different code paths; the similarity is only in the vague title and test assertions." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper tokenizer/test cleanup and FP8 dequant logic are unrelated changes in different subsystems." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Whisper decode tests and generation/watermarking validation are different features with no shared bug." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46922", + "accept": false, + "reason": "Whisper test changes and YOLOS device-map fixes are unrelated model-specific issues." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46539", + "accept": false, + "reason": "These are both test/integration PRs, but they concern different models and different failure modes." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "CLIP/ViT test edits do not overlap with the FP8 loading/dequant fix." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "Model test scaffolding changes are not the same as the generation watermarking validation fix." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46922", + "accept": false, + "reason": "Vague test fixes for CLIP/ViT are not the same underlying bug as YOLOS device placement." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46539", + "accept": false, + "reason": "FP8 dequantization/loading code and Phimoe CI integration setup are different problems." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "These modify unrelated areas: fine-grained FP8 quantization versus generation config validation." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46922", + "accept": false, + "reason": "No shared code path: FP8 kernel/loading changes versus YOLOS device-map behavior." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46539", + "accept": false, + "reason": "Generation watermarking validation and Phimoe CI memory/device setup are unrelated." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46922", + "accept": false, + "reason": "Different models and different bugs: generation config validation versus YOLOS device mismatch." + }, + { + "left": "pull_request:46539", + "right": "pull_request:46922", + "accept": false, + "reason": "Phimoe CI test-environment changes do not match YOLOS runtime device placement fixes." + }, + { + "left": "pull_request:46886", + "right": "pull_request:46922", + "accept": true, + "reason": "Both target the same YOLOS device_map=\"auto\" mismatch issue in the same file and would plausibly be merged into one PR; 46922 is the more complete version." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The 46886/46922 pair is well supported as the same YOLOS bugfix family: both target the same issue, touch the same file, and 46922 is a superset with the forward device-placement fix plus the _no_split_modules guard. The other soft-edge rejections also look appropriately cautious and not overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 8, + "component_count": 33, + "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": "25fc486194bee22a0ae809ed7a7b351a555757657573a307889a8bc27281686e", + "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 open \"don't merge\" PRs that are only superficially similar but modify different workflows/code paths and are not duplicates of one another.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single PR cleanly represents the cluster: 43488 is a packaging/version-bump test PR, 43532 rewires several CI/commenting utilities, and 46747 changes a security-gate workflow trigger.", + "best_issue_reason": null, + "best_pr_reason": "No best PR for dedupe purposes; none of these appear to fix the same underlying bug or change, so there is no safe canonical representative.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:46747", + "accept": false, + "reason": "Both are test/don't-merge PRs, but they touch unrelated files and different behaviors: package metadata/dependency setup vs. a workflow trigger change. No shared concrete bug or mergeable code-path fix." + }, + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Only a superficial overlap in a Vit test file; the actual changes are unrelated (setup/version/dependency changes vs. CI workflow and utility rewrites). Not the same underlying change." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46747", + "accept": false, + "reason": "These modify different CI workflows and code areas, with no evidence they address the same bug or would be merged into one PR. Similarity is broad subsystem overlap only." + } + ] + }, + "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": 33, + "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": "75ca5553ab98c59b121fedc23f4b97c2ac4216625144e734437d950c0421a984", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two separate feature PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They overlap on shared documentation/registration scaffolding, but the actual pipelines, model integrations, and test coverage are different, so they are not duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for a duplicate cluster: the PRs implement different pipelines and touch different underlying model code paths.", + "best_issue_reason": null, + "best_pr_reason": "PR 43613 is the broader representative of the group, with more model integrations and test coverage, but it should not be treated as the same change as 43612.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different features and code paths: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation with different model backends and tests. Shared docs/registration files are incidental scaffolding, not evidence of 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": 10, + "component_count": 33, + "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": "7e89e940fbcbd9f05b9636714e2cd8f9baad3a4e48eac6bc24eb8b9ddf47ab1a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4291, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4291, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs both modify `trainer.py` but address different trainer extension behaviors: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom/PEFT trainers. They are not the same underlying fix.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the changes target separate code paths and solve different problems, so neither PR is a duplicate of the other.", + "best_issue_reason": null, + "best_pr_reason": "PR #43651 is the more concrete bugfix and easier representative of the cluster, but it is still not a duplicate of #43663.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Different problems in the same file: loss scaling for gradient accumulation vs signature/column filtering. They 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": 11, + "component_count": 33, + "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": "c30b4fc567f3b5071d1e63549bae4a7000b89b6f49dc459651f4cdde0560c577", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch the same CLI serve file and both address TypeAdapter/annotation-related NameErrors, but they appear to fix different concrete cases and use different mechanisms. I would not merge them as duplicates.", + "confidence": 0.73, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43656 is the broader, earlier fix with more discussion and a larger typing/import cleanup in the same code path, so it is the better canonical representative of this cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR 43656 is the strongest standalone PR here because it appears to address the core CLI annotation NameError more comprehensively, while 43842 is a narrower follow-up aimed at the optional pydantic-import case.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Both are in `src/transformers/cli/serve.py` and both mention TypeAdapter/annotations, but the diffs point to different fixes: 43656 is a broader future-annotations/typing cleanup, while 43842 specifically handles the `pydantic`-missing case by guarding the import and quoting the annotation. They do not look like the same concrete bug 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": 12, + "component_count": 33, + "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": 33, + "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:44024", + "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:44018|pull_request:44066", + "pull_request:44066|pull_request:44068", + "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:44018|pull_request:44116", + "pull_request:44068|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44068|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44098|pull_request:44101", + "pull_request:44059|pull_request:44086", + "pull_request:44018|pull_request:44071", + "pull_request:44059|pull_request:44722", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44073", + "pull_request:44017|pull_request:44722", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44017|pull_request:44071", + "pull_request:44071|pull_request:44086", + "pull_request:44073|pull_request:44161", + "pull_request:44073|pull_request:44101", + "pull_request:44059|pull_request:44071", + "pull_request:44018|pull_request:44059", + "pull_request:44056|pull_request:44101", + "pull_request:44059|pull_request:44066", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44017", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44018", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request: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:43997|pull_request:44129", + "pull_request:44071|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", + "pull_request:44024|pull_request:44076", + "pull_request:44073|pull_request:44074" + ], + "prepared_review_unit_hash": "f90230f90790d39131cc31d2133b3286ea6618ee22b71ca74f0fc5504ba70c5d", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 76, + "serialized_chars": 84117, + "estimated_input_tokens": 21030, + "estimated_eval_tokens": 42316 + }, + "budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 72, + "serialized_chars": 76938, + "estimated_input_tokens": 19235, + "estimated_eval_tokens": 38726 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of model-specific refactors to standardize output tracing with capture_outputs/can_return_tuple. They share a common umbrella tracker (#43979), but most are not true duplicates because they touch different models and code paths. The only strong duplicate-like pairs are same-model revisions for GPT-Neo and GPT-J/CodeGen.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #43979 is the shared umbrella tracker referenced by many PRs and best matches the cluster\u2019s common output-tracing refactor theme.", + "canonical_pr_reason": "PR #44722 is the broadest and most representative PR: it covers the core GPT-J/CodeGen code paths, explicitly ties to #43979, and looks like the most complete revision of that change.", + "best_issue_reason": "No standalone issue artifact is present in the cluster; #43979 is still the closest canonical tracker for the shared refactor series.", + "best_pr_reason": "PR #44722 is the strongest standalone representative because it subsumes narrower GPT-J-only work and most clearly captures the shared change pattern.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model code paths (CVT/FNet vs GPT-J); same output-tracing theme is too broad." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "Different models and implementations; only the refactor pattern is shared." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Same GPT-J output-tracing refactor on the same code path; the GPT-J-only PR looks like a narrower version of the GPT-J/CodeGen change." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "Different models (RegNet/ResNet vs EfficientNet); this is a shared refactor pattern, not the same bug/change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet vs GPT-J/CodeGen are unrelated concrete changes despite the shared standardization theme." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "Different model families and code paths; not the same underlying change." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are unrelated model-specific refactors." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen versus MGP-STR are different concrete code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen and MPT are separate model implementations; same refactor style is not enough." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo model and same output-tracing refactor; this looks like a later revision of the same concrete change." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo versus GPT-J/CodeGen are different model paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "Different models and files; only the standardized output-tracing theme is shared." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet is unrelated to GPT-J/CodeGen; same umbrella refactor pattern only." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are different concrete code paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are unrelated model-specific changes." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are different underlying changes." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are different concrete code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated model-specific refactors." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are distinct models and changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are different code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo and GPT-J are different model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT and GPT-J are different concrete code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated model-specific refactors." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are different models and code paths." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert are separate model implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-2 and MGP-STR are unrelated concrete changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "Segformer and LongT5 are different model families; only the refactor pattern matches." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "Segformer and VisualBERT are unrelated model-specific changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "Segformer and GPT-J/CodeGen are different code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are different models." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBERT and TextNet are different concrete model implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-2 and MPT are separate model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44059", + "accept": false, + "reason": "GPT-Neo and GPT-2 are different model-specific changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44101", + "accept": false, + "reason": "MPNet and Flaubert are unrelated 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 model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are different concrete changes." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert are unrelated model-specific changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are different models and code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are different concrete changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT and MGP-STR are different code paths." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44161", + "accept": false, + "reason": "VisualBERT and LongT5 are unrelated model-specific changes." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44101", + "accept": false, + "reason": "VisualBERT and Flaubert are different models." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-2 and GPT-J/CodeGen are different model code paths." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", + "accept": false, + "reason": "Flaubert and LongT5 are unrelated model implementations." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "CVT/FNet and Segformer are different concrete changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT and DeBERTa are different model code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are unrelated model-specific refactors." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are different models; same tracing pattern only." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBERT and TextNet are different model-specific changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44018", + "accept": false, + "reason": "SqueezeBERT and GPT-Neo are unrelated concrete 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 unrelated model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT and GPT-J are different concrete changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are unrelated model-specific changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "FNet/CVT and LongT5 are different code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: the cluster is a set of model-specific output-tracing refactors, #43979 is the shared umbrella issue target, and the chosen canonical PR (#44722) is plausibly the broadest GPT-J/CodeGen representative. The soft-edge accepts are conservative and supported by same-model overlap, with no clear evidence of overstating duplicate status." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 33, + "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:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44129" + ], + "soft_pairs": [ + "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": "4adf6a220c68e8312df275a89dcbb287852c228b0c481cbaf016fd462fa12ed8", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 76, + "serialized_chars": 84117, + "estimated_input_tokens": 21030, + "estimated_eval_tokens": 42316 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 4, + "serialized_chars": 12757, + "estimated_input_tokens": 3190, + "estimated_eval_tokens": 6636 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a set of model-specific PRs that all migrate different architectures to standardized output-capture patterns, but they are not duplicates of one another. Each touches a different model implementation and would not plausibly be merged into a single fix PR.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44129 is the most representative only because it has an explicit issue target and a more complete refactor, but it is still distinct from the other PRs and not a duplicate of them.", + "best_issue_reason": null, + "best_pr_reason": "PR 44129 is the strongest standalone representative of the cluster: it has an explicit tracked issue, a concrete code-path change in SpeechT5, and the broadest refactor context among the candidates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "Same high-level migration theme, but MPNet and VisualBert are different code paths with model-specific tracing setup. Not the same underlying change and not mergeable as one PR." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "Both are output-tracing refactors, but they target different models and different implementations. Similar subsystem work is not enough to treat them as duplicates." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are unrelated model implementations. The shared decorator migration pattern does not indicate the same bug or change." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "These are separate model-level refactors in different code paths. The explicit issue link in SpeechT5 does not make it a duplicate of the SqueezeBert 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": 14, + "component_count": 33, + "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": "309cc696164f51c127e040b5453baa33d7656d9d9b3016954330d6fa1e7da200", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40185, + "estimated_input_tokens": 10047, + "estimated_eval_tokens": 20350 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40185, + "estimated_input_tokens": 10047, + "estimated_eval_tokens": 20350 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Loose thematic cluster of model-specific output-tracing refactors. The PRs share a similar modernization pattern, but they target different models and code paths, so none look like true duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true canonical duplicate stands out: each PR is a separate model-specific output-tracing refactor, not the same concrete change. If one must be named, 44025 is the clearest representative because it applies the pattern across two related depth-anything files.", + "best_issue_reason": null, + "best_pr_reason": "PR 44025 is the best representative of the cluster\u2019s theme: it is a substantial, concrete output-tracing refactor and touches two closely related files, making it the most central exemplar among these separate model-specific changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations (timm_backbone vs upernet); same refactor style, but not the same underlying change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Different models and forward paths; both are output-tracing cleanups, but not a duplicate bug/change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Separate model modules (timm_backbone vs univnet); similar cleanup pattern only." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Model-specific refactors in different files; no shared concrete code-path problem." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "timm_backbone and falcon_mamba/mamba are unrelated implementations; same theme, not same change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different architectures and files; not the same fixable code-path issue." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Separate model-specific tracing updates; no evidence they are the same underlying bug." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "mobilenet_v1 and univnet are different model code paths; similar refactor, not duplicate." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different models and different forward semantics; not one mergeable duplicate fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Both are output-tracing refactors, but in unrelated model families and separate files." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "mobilenet_v1 vs vision_encoder_decoder; shared modernization pattern only." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model wrappers and code paths; not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "vision_text_dual_encoder and univnet are unrelated models; same refactor theme, not duplicate." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations with separate forward logic; not the same bug/change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "vision_text_dual_encoder vs falcon_mamba/mamba are distinct code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different architectures and files; too broad a similarity to count as duplicate." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Same output-tracing motif, but applied to unrelated model families." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "UnivNet and UperNet are separate models; no shared concrete issue." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model code paths; not plausibly one merged PR." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Separate model-specific tracing refactors; not the same underlying change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44026", + "accept": false, + "reason": "UnivNet vs vision_encoder_decoder; similar pattern only." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "UperNet and falcon_mamba/mamba touch unrelated implementations." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families and code paths; no duplicate fix here." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "falcon_mamba/mamba vs superpoint are unrelated model changes." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "CodeGen and Mamba/Falcon Mamba are distinct implementations; not the same concrete change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Depth-Anything and vision_encoder_decoder are different model paths; same cleanup style only." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder-decoder-style modules, but separate code paths and changes." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Depth-anything vs superpoint; not the same underlying bug or refactor target." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different models (depth-anything vs rwkv); similarity is only at the refactor theme level." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Depth-anything and DPR are unrelated model implementations." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Both are encoder-decoder wrappers, but one is vision and the other speech; distinct concrete code paths." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "vision_encoder_decoder and superpoint are unrelated models; not mergeable as one fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and separate forward wrappers; not duplicates." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "vision_encoder_decoder vs DPR are separate changes with only a broad theme in common." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "speech_encoder_decoder and superpoint are unrelated code paths." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Different models and different output semantics; not the same fix." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "speech_encoder_decoder and DPR are distinct model implementations." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "SuperPoint and RWKV are unrelated model families; same style of refactor only." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "SuperPoint and DPR are different modules with no shared concrete bug." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "RWKV and DPR are separate model code paths; not a duplicate 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": 15, + "component_count": 33, + "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": "c2412163340ff30c8bb3f364e3c18c3a443c6c4fe2a68efcf255afff434f37f4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate model-integration PRs: one adds PenguinVL, the other adds Voxtral TTS. They share standard Transformers scaffolding files, but the model codepaths, configs, processors, and tests are distinct, so they should not be treated as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs implement different models and fix different feature gaps, so there is no single underlying change to merge them under.", + "best_issue_reason": null, + "best_pr_reason": "PR 45401 is the better standalone representative because it is tied to an explicit tracking issue and has a narrower, concrete scope around Voxtral TTS support.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Reject: PenguinVL and Voxtral TTS are unrelated model additions. The overlap is mostly boilerplate integration files (docs and auto registries), while the model implementations, processors, and tests are for different architectures and use cases." + } + ] + }, + "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": 33, + "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": "886ddc9fc4732cfb2e67f525ae3d903a054986dc3a48a4a2e55a9a84ea80e48d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4407, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4407, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are about model/module fusion, but they target different layers: #44979 introduces a generic Module Fusion API, while #45363 adds KernelConfig-based n-to-1 kernel fusion and integration plumbing. They are related, but not the same underlying change and not mergeable as one duplicate PR.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45363 is the more concrete, user-facing fusion feature with broader integration changes; #44979 is a separate generic API layer rather than the same fix/change.", + "best_issue_reason": null, + "best_pr_reason": "#45363 is the strongest representative of the cluster theme because it implements an end-to-end fusion path across multiple files, whereas #44979 is a distinct abstraction/API proposal.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Same broad fusion area, but different abstractions and code paths: generic module fusion API vs KernelConfig-based kernel fusion. Not the same concrete change and not a single plausible merged 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": 17, + "component_count": 33, + "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": "67270ed2190a8f0b83f02d0adc622c2ee4ad6fddec79692b54ba8c1132b0c361", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same AttributeError in `_patch_mistral_regex` by changing the pre-tokenizer access path; PR 45317 is the more complete variant because it adds tests.", + "confidence": 0.98, + "canonical_issue_reason": "Both PRs explicitly target issue #45081, which appears to be the shared bug report for this exact failure.", + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same code fix as 45086 plus test coverage, making it the more complete implementation.", + "best_issue_reason": "No issue item is present in the cluster, but issue #45081 is the common upstream bug being fixed by both PRs.", + "best_pr_reason": "PR 45317 is the best PR candidate since it fixes the same concrete code path as 45086 and adds validation tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying AttributeError fix in the same function and same file; 45317 is essentially the same change with added tests, so these are duplicates/same change family." + } + ] + }, + "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": 33, + "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": "35aeec7833ce9e3634261a4569d530605fe82c6a498007d7a4ab9d0f88a7a5e8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6065, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6065, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are related to NemotronH but they are not duplicates: one is a narrow conversion-mapping / inheritance refactor, while the other is a broader split into NemotronHDense and NemotronHSparse with docs and auto-mapping updates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "If a representative PR must be chosen, 45569 is the broader, more central architectural change; 45115 is a narrower follow-up/refactor focused on checkpoint conversion mappings.", + "best_issue_reason": null, + "best_pr_reason": "45569 is the better representative PR for the cluster because it covers the substantive model-family split and public-facing integration work, whereas 45115 is limited to rename/mapping adjustments.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Different underlying changes: 45115 updates NemotronH conversion mapping and compatibility renames; 45569 introduces separate dense/sparse NemotronH variants plus docs, auto-mapping, and tests. Shared files are incidental to the same model family, not evidence of 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": 19, + "component_count": 33, + "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": "8ac2f336b92588b14df661a0ac7abeef715a3cc9b9295ea8c9e26d16fa0a356d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are about heterogeneous model support, but they are not the same change: #45333 is a focused per-layer configuration PR, while #45332 is a much broader end-to-end heterogeneous modeling feature that also touches cache, masking, integrations, and model wiring. They look related and likely sequential, not duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45332 is the broader umbrella change and best represents the cluster because it includes the configuration work plus the actual heterogeneous modeling/integration implementation.", + "best_issue_reason": null, + "best_pr_reason": "#45332 is the strongest standalone PR for this cluster: it covers the full feature surface and is more likely to be the principal implementation users care about.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Not the same underlying PR: #45333 is config-only, while #45332 adds a larger heterogeneous modeling stack across cache/masking/integrations. They share some files and theme, but they are not mergeable as one duplicate 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": 33, + "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": "cddcef27f5dcbaa652cb0a3b9cb445425898798001ec5cf8578e8b3af4768f56", + "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": "The two PRs are related by typing work and a few shared files, but they address different scopes: one is a broad type-checking sweep across many `src/transformers/*` files, while the other adds targeted typing helpers/protocols and narrower model-utils integration. They do not look like the same underlying change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45425 is the more coherent representative of the cluster: it introduces focused typing abstractions and concrete typing support around `modeling_utils`/related helpers, rather than a wide mechanical type-checking pass.", + "best_issue_reason": null, + "best_pr_reason": "PR 45425 is the best single representative because it is narrower and more semantically focused, making it easier to understand as a distinct change than the broad repo-wide typing cleanup in PR 45415.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Both are typing-related and share a few files, but they solve different problems: 45415 is a broad type-checking enablement pass across many modules, while 45425 adds specific typing protocols/aliases and targeted fixes. They are not the same concrete bug/change and would not naturally merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 33, + "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": "9c51e3f56fb496d68dfa9de2179c5c31038f551b081dcf012d5295582991c893", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are related by expert-parallel loading, but they fix different bugs: one is about preserving/reading `distributed_config` during Accelerate/ZeRO-3 loading, while the other fixes FSDP2 expert parameters being overwritten after sharding. They are adjacent, not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45662 is the better canonical representative for the cluster because it addresses a concrete, self-contained loading bug in the parameter assignment path; PR #45548 is a separate configuration/serialization fix and does not collapse into the same change.", + "best_issue_reason": null, + "best_pr_reason": "PR #45662 is the strongest standalone fix: it targets a specific code path (`set_param_for_module` / post-shard wrapping) and has a clear, reproducible failure mode. PR #45548 is related but covers a different EP/DeepSpeed configuration issue.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Both are EP-loading related, but they fix different problems in different paths: #45548 preserves `distributed_config` and adds `has_ep`, while #45662 prevents expert tensors from being overwritten after sharding via `post_shard_wrap`. 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": "pull_request", + "component_index": 22, + "component_count": 33, + "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": "783efdaa4c40f41bfd54ab812a92843dcf7ba2065e62e98bd9608e67984297a3", + "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 pull requests target the same xdist-safe captured_info logging problem under issue #45561; 45645 is a broader superset of 45639 that also updates CI log collection.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45645 is the better canonical representative because it includes the core testing_utils xdist-safe fix plus the CI workflow change needed to preserve and collect the split log files.", + "best_issue_reason": null, + "best_pr_reason": "PR #45645 is the strongest overall match for the cluster: it addresses the same bug as #45639 and adds the downstream artifact-handling changes needed to make the fix complete in CI.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "They fix the same underlying captured_info/xdist collision issue in the same code path, and #45645 plausibly subsumes #45639 with additional workflow support." + } + ] + }, + "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": 33, + "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": 24, + "component_count": 33, + "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": "2ca6a640f3586da2b050586c5342f5ec0325257d090037859fb0a09d5c2d73fb", + "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": "The two pull requests overlap on distributed code, but they are not the same change: one is a focused EP/trainer integration feature, while the other is a broad distributed branch base spanning many unrelated files. The soft duplicate edge should be rejected.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46126 is the more specific, self-contained change set: it targets EP routing behavior and trainer integration on top of DistributedConfig, with a narrow set of touched files and a clear feature intent. PR #46269 is much broader and looks like a base/rollup branch rather than a single concrete fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #46126 is the best representative PR for the cluster because it has the clearest single-purpose change and the tightest scope. PR #46269 is too wide-ranging to serve as the canonical example for this cluster.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Reject: they share distributed filenames, but the change intent is different. #46126 is an EP + trainer integration feature with router-parallel logic; #46269 is a broad distributed branch base touching many subsystems, so they do not look like the same underlying PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 25, + "component_count": 33, + "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": "0d9e6be9906837afab8451f5bac450132ba2805572d2cd101d36b79e6d816443", + "component_budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26665, + "estimated_input_tokens": 6667, + "estimated_eval_tokens": 13590 + }, + "budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26665, + "estimated_input_tokens": 6667, + "estimated_eval_tokens": 13590 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are a broad set of workflow-maintenance edits, not true duplicates. The only tight cluster is several changes to pr-ci-security-gate.yml, but they modify different parts of that workflow and should not be merged as one duplicate change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear canonical PR: the set splits into unrelated workflow tweaks, and even the pr-ci-security-gate edits are distinct changes to the same file rather than one duplicate fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 46554 is the most substantive and self-contained change in the security-gate subgroup, but it is still not a duplicate of the later permissions/reusable-workflow/secrets-only updates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46138", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflow files and different behavior; both are CI chores, but they do not fix the same underlying change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46558", + "accept": false, + "reason": "Same broad CI area, but one changes self-comment message piping while the other adjusts PR security-gate permissions and approval flow." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46563", + "accept": false, + "reason": "Unrelated workflow changes; shared CI-maintenance style is not enough for duplication." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46716", + "accept": false, + "reason": "Different workflow and different concern; one updates comment handling, the other tightens security-gate secrets." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Both are shell-safety tweaks, but they affect different workflows and different code paths." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "Different workflows and unrelated runtime behavior; no shared underlying bug or fix." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "One updates ssh-runner setup, the other changes anti-slop permissions; not the same change." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Different workflow files and unrelated logic." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46554", + "accept": false, + "reason": "SSH runner hardening is unrelated to pr-ci security-gate logic." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflow concerns; one manages SSH runner env, the other modifies PR CI security-gate behavior." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Both touch workflows, but they change different jobs and unrelated behavior." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Different workflow files with unrelated fixes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "No same underlying bug/change; permissions tweak vs shell input handling in a separate workflow." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflows and different fixes." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Both are workflow chores, but they address different workflows and behaviors." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Different files and different logic; not a duplicate fix." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46554", + "accept": false, + "reason": "Unrelated workflow changes." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflow files and different behavior; no common underlying change." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46554", + "accept": false, + "reason": "Benchmark dispatch support is unrelated to PR security-gate changes." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46554", + "accept": false, + "reason": "Similar shell-hardening theme, but they affect different workflows and code paths." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflows and distinct fixes." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46558", + "accept": false, + "reason": "Same file, but different underlying changes: one adds security checks, the other changes workflow permissions/approval gating." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46563", + "accept": false, + "reason": "Both touch pr-ci-security-gate.yml, but they modify different execution models and are not the same fix." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46716", + "accept": false, + "reason": "Same workflow file, but one adds code-level checks while the other only changes secrets handling." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46563", + "accept": false, + "reason": "Different edits to the same workflow file; not the same underlying bug/change." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46716", + "accept": false, + "reason": "Distinct security-gate tweaks: permissions/approval logic vs secrets inheritance." + }, + { + "left": "pull_request:46563", + "right": "pull_request:46716", + "accept": false, + "reason": "Different changes in the same workflow file, but not one duplicate fix." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46563", + "accept": false, + "reason": "Different workflows and unrelated behavior." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46563", + "accept": false, + "reason": "SSH runner hardening and PR security-gate documentation are unrelated." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46716", + "accept": false, + "reason": "Different workflow files and different concerns." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46716", + "accept": false, + "reason": "Unrelated workflow changes; no shared code-path problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 26, + "component_count": 33, + "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": 27, + "component_count": 33, + "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:47162" + ], + "soft_pairs": [ + "pull_request:46310|pull_request:47162" + ], + "prepared_review_unit_hash": "b6c0f043a84c4fe76d1bab1c61daf066cbfad2d7e8c98554373caccc0cb4bfcf", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4883, + "estimated_input_tokens": 1221, + "estimated_eval_tokens": 2698 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4883, + "estimated_input_tokens": 1221, + "estimated_eval_tokens": 2698 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs fix the same off-by-one error in the MoonshineStreaming sliding window right bound; PR 47162 is the more complete follow-up because it includes the Moonshine fix plus tests and an additional Gemma4 instance of the same bug.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47162 is the best canonical PR because it contains the same MoonshineStreaming fix as PR 46310, adds tests, and appears to be the more complete version of the patch set.", + "best_issue_reason": null, + "best_pr_reason": "PR 47162 is the strongest representative of the cluster: it fixes the same concrete off-by-one bug in MoonshineStreaming and also validates the behavior with tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46310", + "right": "pull_request:47162", + "accept": true, + "reason": "Both PRs correct the same right-window off-by-one in sliding_window_mask_function for MoonshineStreaming; 47162 is a superset rather than a different change, so they are close enough to treat as the same underlying 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": 28, + "component_count": 33, + "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": "20719eedad56e6d08f87c43bb1395627a2c86dce7f938b1f206eef095ce5bde7", + "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 both tokenizer-related PRs, but they address different problems in different code paths: CodeLlama decode stripping real leading spaces vs. DeepSeek checkpoint-to-tokenizer-class selection in auto-loading.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR stands out because the two PRs fix unrelated tokenizer bugs in separate modules and could not plausibly be merged into one change.", + "best_issue_reason": null, + "best_pr_reason": "Neither PR is a good global representative for the cluster since there is no shared underlying bug or change; each is only representative of its own distinct fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46574", + "right": "pull_request:46580", + "accept": false, + "reason": "Different failure modes and code paths: 46574 changes CodeLlama decode behavior in a model-specific tokenizer file, while 46580 changes auto tokenizer class resolution for specific DeepSeek checkpoints. They do not look like the same underlying bug or a mergeable single 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": 29, + "component_count": 33, + "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-46696-3", + "nodes": [ + "pull_request:46696", + "pull_request:47134", + "pull_request:47215" + ], + "soft_pairs": [ + "pull_request:46696|pull_request:47215", + "pull_request:47134|pull_request:47215", + "pull_request:46696|pull_request:47134" + ], + "prepared_review_unit_hash": "bb884b30e85cd2ceef95eede3c27b196fab906fe003e0b230d86edbd296caeea", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7370, + "estimated_input_tokens": 1843, + "estimated_eval_tokens": 3942 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7370, + "estimated_input_tokens": 1843, + "estimated_eval_tokens": 3942 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are all about flash-attention long-context performance regressions, but they split into two concrete code paths: the unpad/max_seqlen scalar handling and the kwargs-processing/tracing path. PR 47215 is the most complete and likely final fix; 47134 is a partial earlier version of the same unpad regression. 46696 is related but addresses a different helper path, so it is not a duplicate of 47134.", + "confidence": 0.69, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47215 is the best canonical PR: it is the latest and most complete fix, uses the compile-aware guard, and includes the broader test/benchmark follow-through for the same flash-attention regression.", + "best_issue_reason": null, + "best_pr_reason": "PR 47215 is the strongest representative because it most fully addresses the regression and appears to subsume the earlier partial fix in 47134.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46696", + "right": "pull_request:47215", + "accept": true, + "reason": "Both target the same flash-attention long-context regression in the same file and issue thread; 47215 looks like the broader/finalized version of the same fix family." + }, + { + "left": "pull_request:47134", + "right": "pull_request:47215", + "accept": true, + "reason": "These patch the same unpad/max_seqlen performance bug in the same helper functions; 47215 is a refined, compile-aware replacement for the earlier partial fix." + }, + { + "left": "pull_request:46696", + "right": "pull_request:47134", + "accept": false, + "reason": "They touch different concrete code paths in flash-attention utils: one fixes kwargs/tracing scalar handling, the other fixes unpad/max_seqlen materialization. Related, but not the same bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. It correctly distinguishes the two concrete flash-attention paths: kwargs/tracing scalar handling in 46696 versus unpad/max_seqlen materialization in 47134/47215. The canonical-choice rationale for 47215 is supported by the later timestamp, broader file coverage, and compile-aware guard. The accepted soft pairs are also plausible mergeable same-bugfix candidates, and the rejected 46696\u219447134 pair is appropriately not overstated as a duplicate." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 30, + "component_count": 33, + "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-46773-2", + "nodes": [ + "pull_request:46773", + "pull_request:47087" + ], + "soft_pairs": [ + "pull_request:46773|pull_request:47087" + ], + "prepared_review_unit_hash": "aee84dd9f4239c8957485bfa12db9c287541b107804d186e6b2993c105c18739", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5433, + "estimated_input_tokens": 1359, + "estimated_eval_tokens": 2974 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5433, + "estimated_input_tokens": 1359, + "estimated_eval_tokens": 2974 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch padding-mask handling in attention-like models, but they fix different bugs: one restores masking for batch size 1 in linear-attention models, while the other fixes recurrent masking during continued forwards/chunked prefill/cache continuation. They are related in theme but not the same underlying change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47087 is the better canonical PR because it has a clear, specific bug statement, an explicit issue target, and a focused fix in masking_utils with tests covering recurrent continued-forward behavior.", + "best_issue_reason": null, + "best_pr_reason": "PR 46773 is also a plausible representative for the linear-attention padding-mask bug, but it is narrower and less directly tied to the recurrent-mask behavior than 47087. For this cluster, 47087 is the strongest overall representative of the masking regression theme.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46773", + "right": "pull_request:47087", + "accept": false, + "reason": "Both concern padding masks, but 46773 fixes a batch-size-1 masking gate in linear-attention model forward paths, while 47087 fixes mask trimming/skipping for continued forwards in recurrent attention. Different code paths and failure modes; 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": 33, + "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-47032-5", + "nodes": [ + "pull_request:47032", + "pull_request:47074", + "pull_request:47076", + "pull_request:47195", + "pull_request:47197" + ], + "soft_pairs": [ + "pull_request:47032|pull_request:47197", + "pull_request:47074|pull_request:47076", + "pull_request:47074|pull_request:47197", + "pull_request:47074|pull_request:47195", + "pull_request:47195|pull_request:47197" + ], + "prepared_review_unit_hash": "5c41ec08009a21b305c11bdf4b96e53c09f8f785828ee07d7f373fb66ac77de4", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 11473, + "estimated_input_tokens": 2869, + "estimated_eval_tokens": 5994 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 11473, + "estimated_input_tokens": 2869, + "estimated_eval_tokens": 5994 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are all transformer multimodal/generation fixes, but they address different concrete bugs in different code paths; none of the soft pairs look like true duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47074 is the best representative if one must be chosen: it fixes a concrete, user-facing multimodal pipeline input-normalization bug and is the broadest of the set.", + "best_issue_reason": null, + "best_pr_reason": "PR 47074 is the strongest cluster representative because it targets the most general pipeline-level behavior and has the clearest, concrete bug fix among the group.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47032", + "right": "pull_request:47197", + "accept": false, + "reason": "Different bugs in different components: stop_sequence handling in image-text-to-text vs batched decode/postprocessing in Qwen Omni." + }, + { + "left": "pull_request:47074", + "right": "pull_request:47076", + "accept": false, + "reason": "Both are multimodal, but one normalizes AnyToAnyPipeline input keys while the other validates Gemma4 audio placeholders; distinct code paths and failure modes." + }, + { + "left": "pull_request:47074", + "right": "pull_request:47195", + "accept": false, + "reason": "Unrelated fixes: multimodal input-key normalization vs generation candidate length-capping in the lookup decoder." + }, + { + "left": "pull_request:47074", + "right": "pull_request:47197", + "accept": false, + "reason": "Separate issues in separate subsystems: AnyToAnyPipeline input mapping vs Qwen Omni text postprocessing." + }, + { + "left": "pull_request:47195", + "right": "pull_request:47197", + "accept": false, + "reason": "Generation candidate selection length cap and Qwen Omni batched decoding are 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": "pull_request", + "component_index": 32, + "component_count": 33, + "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-47105-2", + "nodes": [ + "pull_request:47105", + "pull_request:47129" + ], + "soft_pairs": [ + "pull_request:47105|pull_request:47129" + ], + "prepared_review_unit_hash": "7e2cd2353573798238b25945542e01d22aa7df063e18b0fe22acae597befdf4e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4181, + "estimated_input_tokens": 1046, + "estimated_eval_tokens": 2348 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4181, + "estimated_input_tokens": 1046, + "estimated_eval_tokens": 2348 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are unrelated: one re-enables a VITS model forward test, while the other adjusts an RT-DETR v2 weight-conversion script for current model keys. They do not fix the same bug or change the same concrete code path.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: PR #47105 is a VITS test update, while PR #47129 is an RT-DETR v2 converter fix; they target different subsystems and problems.", + "best_issue_reason": null, + "best_pr_reason": "No single best PR for deduplication. Both are valid standalone changes, but neither is a duplicate of the other or a better representative for the pair.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47105", + "right": "pull_request:47129", + "accept": false, + "reason": "Different models and code paths: VITS test re-enable vs RT-DETR v2 converter key mapping 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": 33, + "component_count": 33, + "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-47122-5", + "nodes": [ + "pull_request:47122", + "pull_request:47123", + "pull_request:47124", + "pull_request:47150", + "pull_request:47224" + ], + "soft_pairs": [ + "pull_request:47122|pull_request:47123", + "pull_request:47122|pull_request:47224", + "pull_request:47122|pull_request:47124", + "pull_request:47122|pull_request:47150", + "pull_request:47123|pull_request:47124", + "pull_request:47123|pull_request:47150", + "pull_request:47124|pull_request:47150", + "pull_request:47150|pull_request:47224" + ], + "prepared_review_unit_hash": "47e98445d8efee7c841090061eb41bc0366a4a1a767495c63ef5f9a73d29062b", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 12423, + "estimated_input_tokens": 3106, + "estimated_eval_tokens": 6468 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 12423, + "estimated_input_tokens": 3106, + "estimated_eval_tokens": 6468 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated PRs for different model/test failures, with one umbrella PR (47224) that appears to subsume at least the EdgeTam and Whisper fixes.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47224 is the broad umbrella change that bundles multiple Serge integration-test fixes and appears to include the concrete EdgeTam and Whisper code-path corrections from smaller PRs.", + "best_issue_reason": null, + "best_pr_reason": "47224 is the best cluster representative because it is the most comprehensive fix and explicitly targets the shared integration-test-failure cleanup, while the other PRs are narrower point fixes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47122", + "right": "pull_request:47123", + "accept": false, + "reason": "Different failures and code paths: EdgeTam forward kwargs handling vs. MiniMax-M3-VL load/offload test setup." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47224", + "accept": true, + "reason": "47224 includes the same EdgeTam fix theme in the same model files (dropping unsupported original_sizes/backbone kwargs), so it looks like the same underlying change." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47124", + "accept": false, + "reason": "EdgeTam kwargs filtering is unrelated to Falcon-Mamba generation expectation updates." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47150", + "accept": false, + "reason": "EdgeTam backbone kwarg cleanup and Whisper max-length handling are separate bugs." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47124", + "accept": false, + "reason": "MiniMax-M3-VL offload-folder test fix is unrelated to Falcon-Mamba output mismatches." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47150", + "accept": false, + "reason": "MiniMax-M3-VL loading issue and Whisper generation-length issue are different problems." + }, + { + "left": "pull_request:47124", + "right": "pull_request:47150", + "accept": false, + "reason": "Both are test-mismatch PRs, but they fix different models and different failure modes." + }, + { + "left": "pull_request:47150", + "right": "pull_request:47224", + "accept": true, + "reason": "47224 appears to include the same Whisper generation_whisper.py correction for max_length on previous-segment conditioning, so this is the same underlying fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The representative PR choice is supported by the umbrella nature of 47224, and the accepted soft-edge pairs (47122\u219447224, 47150\u219447224) are plausibly mergeable as the same bugfix themes. The rejected pairs are also reasonably treated as unrelated point fixes across different models/failures." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/manifest.json b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..a3970a1f23f094ffecaf775f378a2111e83244d2 --- /dev/null +++ b/snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260710t000031z", + "artifacts": { + "hybrid": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-07-10T00:11:43Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260710T000031Z", + "variant": "hybrid" +} diff --git a/snapshots/20260710T000031Z/manifest.json b/snapshots/20260710T000031Z/manifest.json index 2ceb7f8dfe407e3a8e1814acba1b25cb4f3066b8..7a69c337f4ff72885f3bcf7b602be766f6c92e97 100644 --- a/snapshots/20260710T000031Z/manifest.json +++ b/snapshots/20260710T000031Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 450 }, "extracted_at": "2026-07-10T00:00:31Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260710t000031z", + "runs": { + "hybrid-model-20260710t000031z": { + "analysis_id": "hybrid-model-20260710t000031z", + "artifacts": { + "hybrid": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260710T000031Z/analysis-runs/hybrid-model-20260710t000031z/manifest.json", + "model": null, + "published_at": "2026-07-10T00:11:43Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260710T000031Z", "watermark": {