diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 402ea479b7ecef1a09f905cfc37e58dba4b6e3ad..e8e2e4fbcd9e644e7114fb81f1a098a0eb8ea0d2 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": "20260705T000031Z", - "generated_at": "2026-07-05T00:05:52Z", + "snapshot_id": "20260705T060758Z", + "generated_at": "2026-07-05T06:18:09Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1050,232 +1050,306 @@ ] }, { - "cluster_id": "cluster-43451-4", - "summary": "Cluster of 4 related pull requests linked by soft_similarity.", + "cluster_id": "cluster-46907-7", + "summary": "Cluster of 7 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 43451, + "canonical_pr_number": 46907, "issue_numbers": [], "pr_numbers": [ - 43451, - 46707, - 46836, - 47003 + 46907, + 46927, + 46949, + 46951, + 46971, + 47055, + 47056 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 43451, - "right_pr_number": 46707, - "code_similarity": 0.026, - "size_similarity": 0.069, + "left_pr_number": 46907, + "right_pr_number": 46927, + "code_similarity": 0.209, + "size_similarity": 0.929, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.079, + "patch_similarity": 0.158, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 43451, - "right_pr_number": 46836, - "code_similarity": 0.261, - "size_similarity": 0.792, - "file_overlap": 0.176, + "left_pr_number": 46907, + "right_pr_number": 46949, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.332, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46907, + "right_pr_number": 46951, + "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": 46907, + "right_pr_number": 46971, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46907, + "right_pr_number": 47055, + "code_similarity": 0.925, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 0.5, "shared_filenames": [ - "docs/source/en/_toctree.yml", - "src/transformers/conversion_mapping.py", - "src/transformers/models/__init__.py", - "src/transformers/models/auto/auto_mappings.py", - "src/transformers/models/auto/modeling_auto.py", - "utils/check_repo.py" + "src/transformers/models/edgetam/modeling_edgetam.py", + "src/transformers/models/edgetam/modular_edgetam.py" ], "shared_file_areas": [ { - "filename": "docs/source/en/_toctree.yml", - "left_ranges": [ - [ - 1367, - 1374 - ] - ], - "right_ranges": [ - [ - 887, - 894 - ] - ] - }, - { - "filename": "src/transformers/conversion_mapping.py", + "filename": "src/transformers/models/edgetam/modeling_edgetam.py", "left_ranges": [ [ - 664, - 696 + 464, + 472 ] ], "right_ranges": [ [ - 1420, - 1486 + 464, + 472 ] ] }, { - "filename": "src/transformers/models/__init__.py", + "filename": "src/transformers/models/edgetam/modular_edgetam.py", "left_ranges": [ [ - 289, - 295 + 210, + 218 ] ], "right_ranges": [ [ - 452, - 458 + 210, + 218 ] ] - }, - { - "filename": "src/transformers/models/auto/auto_mappings.py", - "left_ranges": [ - [ - 386, - 393 - ], - [ - 816, - 822 - ], - [ - 936, - 942 - ], - [ - 1057, - 1063 - ], - [ - 1189, - 1195 - ] - ], - "right_ranges": [ - [ - 628, - 638 - ], - [ - 905, - 914 - ], - [ - 1112, - 1118 - ], - [ - 1238, - 1244 - ] - ] - }, - { - "filename": "src/transformers/models/auto/modeling_auto.py", - "left_ranges": [ - [ - 326, - 333 - ], - [ - 1087, - 1093 - ], - [ - 1906, - 1912 - ] - ], - "right_ranges": [ - [ - 501, - 507 - ], - [ - 1107, - 1113 - ] - ] - }, + } + ] + }, + { + "left_pr_number": 46907, + "right_pr_number": 47056, + "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": 46927, + "right_pr_number": 46949, + "code_similarity": 0.157, + "size_similarity": 0.786, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 46951, + "code_similarity": 0.079, + "size_similarity": 0.393, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 46971, + "code_similarity": 0.162, + "size_similarity": 0.786, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 47055, + "code_similarity": 0.209, + "size_similarity": 0.929, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.158, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 47056, + "code_similarity": 0.079, + "size_similarity": 0.393, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 46951, + "code_similarity": 0.1, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 46971, + "code_similarity": 0.206, + "size_similarity": 1.0, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 47055, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 47056, + "code_similarity": 0.54, + "size_similarity": 0.5, + "file_overlap": 0.5, + "area_overlap": 0.4, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/whisper/modeling_whisper.py" + ], + "shared_file_areas": [ { - "filename": "utils/check_repo.py", + "filename": "src/transformers/models/whisper/modeling_whisper.py", "left_ranges": [ [ - 250, - 258 - ], - [ - 333, - 340 + 309, + 315 ] ], "right_ranges": [ [ - 267, - 274 - ], - [ - 489, - 496 + 306, + 312 ] ] } ] }, { - "left_pr_number": 43451, - "right_pr_number": 47003, - "code_similarity": 0.174, - "size_similarity": 0.738, + "left_pr_number": 46951, + "right_pr_number": 46971, + "code_similarity": 0.106, + "size_similarity": 0.5, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.178, + "patch_similarity": 0.038, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 46707, - "right_pr_number": 46836, - "code_similarity": 0.029, - "size_similarity": 0.087, + "left_pr_number": 46951, + "right_pr_number": 47055, + "code_similarity": 0.085, + "size_similarity": 0.423, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.08, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 46707, - "right_pr_number": 47003, - "code_similarity": 0.028, - "size_similarity": 0.094, + "left_pr_number": 46951, + "right_pr_number": 47056, + "code_similarity": 0.2, + "size_similarity": 1.0, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.061, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 46836, - "right_pr_number": 47003, - "code_similarity": 0.214, - "size_similarity": 0.932, + "left_pr_number": 46971, + "right_pr_number": 47055, + "code_similarity": 0.169, + "size_similarity": 0.846, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.183, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46971, + "right_pr_number": 47056, + "code_similarity": 0.106, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47055, + "right_pr_number": 47056, + "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": [] } @@ -1365,14 +1439,189 @@ "filename": "tests/models/deimv2/test_modeling_deimv2.py", "left_ranges": [ [ - 1, - 15 + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 1734 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-43451-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 43451, + "issue_numbers": [], + "pr_numbers": [ + 43451, + 46836 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43451, + "right_pr_number": 46836, + "code_similarity": 0.261, + "size_similarity": 0.792, + "file_overlap": 0.176, + "area_overlap": 0.0, + "patch_similarity": 0.332, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "src/transformers/conversion_mapping.py", + "src/transformers/models/__init__.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py", + "utils/check_repo.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 1367, + 1374 + ] + ], + "right_ranges": [ + [ + 887, + 894 + ] + ] + }, + { + "filename": "src/transformers/conversion_mapping.py", + "left_ranges": [ + [ + 664, + 696 + ] + ], + "right_ranges": [ + [ + 1420, + 1486 + ] + ] + }, + { + "filename": "src/transformers/models/__init__.py", + "left_ranges": [ + [ + 289, + 295 + ] + ], + "right_ranges": [ + [ + 452, + 458 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 386, + 393 + ], + [ + 816, + 822 + ], + [ + 936, + 942 + ], + [ + 1057, + 1063 + ], + [ + 1189, + 1195 + ] + ], + "right_ranges": [ + [ + 628, + 638 + ], + [ + 905, + 914 + ], + [ + 1112, + 1118 + ], + [ + 1238, + 1244 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 326, + 333 + ], + [ + 1087, + 1093 + ], + [ + 1906, + 1912 + ] + ], + "right_ranges": [ + [ + 501, + 507 + ], + [ + 1107, + 1113 + ] + ] + }, + { + "filename": "utils/check_repo.py", + "left_ranges": [ + [ + 250, + 258 + ], + [ + 333, + 340 ] ], "right_ranges": [ [ - 1, - 1734 + 267, + 274 + ], + [ + 489, + 496 ] ] } @@ -1381,41 +1630,112 @@ ] }, { - "cluster_id": "cluster-46926-6", - "summary": "Cluster of 6 related pull requests linked by soft_similarity.", + "cluster_id": "cluster-46908-5", + "summary": "Cluster of 5 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 46927, + "canonical_pr_number": 46926, "issue_numbers": [], "pr_numbers": [ + 46908, 46926, - 46927, - 46949, 47054, - 47055, - 47056 + 47057, + 47058 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46926, - "right_pr_number": 46927, - "code_similarity": 0.086, - "size_similarity": 0.429, + "left_pr_number": 46908, + "right_pr_number": 46926, + "code_similarity": 0.31, + "size_similarity": 0.8, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.0, + "patch_similarity": 1.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 46926, - "right_pr_number": 46949, - "code_similarity": 0.109, - "size_similarity": 0.545, + "left_pr_number": 46908, + "right_pr_number": 47054, + "code_similarity": 0.31, + "size_similarity": 0.8, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 1.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46908, + "right_pr_number": 47057, + "code_similarity": 0.565, + "size_similarity": 0.231, + "file_overlap": 1.0, + "area_overlap": 0.565, + "patch_similarity": 0.143, + "shared_filenames": [ + "tests/models/deepseek_v32/test_modeling_deepseek_v32.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/deepseek_v32/test_modeling_deepseek_v32.py", + "left_ranges": [ + [ + 13, + 19 + ], + [ + 251, + 257 + ], + [ + 270, + 276 + ], + [ + 291, + 297 + ], + [ + 332, + 338 + ] + ], + "right_ranges": [ + [ + 13, + 19 + ], + [ + 248, + 260 + ], + [ + 268, + 281 + ], + [ + 290, + 303 + ], + [ + 332, + 345 + ] + ] + } + ] + }, + { + "left_pr_number": 46908, + "right_pr_number": 47058, + "code_similarity": 0.15, + "size_similarity": 0.75, "file_overlap": 0.0, "area_overlap": 0.0, "patch_similarity": 0.0, @@ -1461,137 +1781,42 @@ }, { "left_pr_number": 46926, - "right_pr_number": 47055, - "code_similarity": 0.092, - "size_similarity": 0.462, + "right_pr_number": 47057, + "code_similarity": 0.058, + "size_similarity": 0.185, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.0, + "patch_similarity": 0.143, "shared_filenames": [], "shared_file_areas": [] }, { "left_pr_number": 46926, - "right_pr_number": 47056, - "code_similarity": 0.183, - "size_similarity": 0.917, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46927, - "right_pr_number": 46949, - "code_similarity": 0.157, - "size_similarity": 0.786, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46927, - "right_pr_number": 47054, - "code_similarity": 0.086, - "size_similarity": 0.429, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46927, - "right_pr_number": 47055, - "code_similarity": 0.209, - "size_similarity": 0.929, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.158, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46927, - "right_pr_number": 47056, - "code_similarity": 0.079, - "size_similarity": 0.393, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46949, - "right_pr_number": 47054, - "code_similarity": 0.109, - "size_similarity": 0.545, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 46949, - "right_pr_number": 47055, - "code_similarity": 0.169, - "size_similarity": 0.846, + "right_pr_number": 47058, + "code_similarity": 0.12, + "size_similarity": 0.6, "file_overlap": 0.0, "area_overlap": 0.0, "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, - { - "left_pr_number": 46949, - "right_pr_number": 47056, - "code_similarity": 0.54, - "size_similarity": 0.5, - "file_overlap": 0.5, - "area_overlap": 0.4, - "patch_similarity": 1.0, - "shared_filenames": [ - "src/transformers/models/whisper/modeling_whisper.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/models/whisper/modeling_whisper.py", - "left_ranges": [ - [ - 309, - 315 - ] - ], - "right_ranges": [ - [ - 306, - 312 - ] - ] - } - ] - }, { "left_pr_number": 47054, - "right_pr_number": 47055, - "code_similarity": 0.092, - "size_similarity": 0.462, + "right_pr_number": 47057, + "code_similarity": 0.058, + "size_similarity": 0.185, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.0, + "patch_similarity": 0.143, "shared_filenames": [], "shared_file_areas": [] }, { "left_pr_number": 47054, - "right_pr_number": 47056, - "code_similarity": 0.183, - "size_similarity": 0.917, + "right_pr_number": 47058, + "code_similarity": 0.12, + "size_similarity": 0.6, "file_overlap": 0.0, "area_overlap": 0.0, "patch_similarity": 0.0, @@ -1599,13 +1824,13 @@ "shared_file_areas": [] }, { - "left_pr_number": 47055, - "right_pr_number": 47056, - "code_similarity": 0.085, - "size_similarity": 0.423, + "left_pr_number": 47057, + "right_pr_number": 47058, + "code_similarity": 0.083, + "size_similarity": 0.308, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.0, + "patch_similarity": 0.143, "shared_filenames": [], "shared_file_areas": [] } @@ -1613,70 +1838,25 @@ }, { "cluster_id": "cluster-46894-4", - "summary": "Cluster of 1 issues and 3 PRs centered on issue #46894.", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46894.", "status": "open", "confidence": 0.8, "canonical_issue_number": 46894, - "canonical_pr_number": 46903, + "canonical_pr_number": 46895, "issue_numbers": [ - 46894 + 46894, + 46897 ], "pr_numbers": [ 46895, - 46903, 46958 ], - "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 - ] - ] - } - ] - }, + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ { "left_pr_number": 46895, "right_pr_number": 46958, @@ -1753,59 +1933,6 @@ ] } ] - }, - { - "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 - ] - ] - } - ] } ] }, @@ -2288,7 +2415,7 @@ "cluster_id": "cluster-45561-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", "status": "open", - "confidence": 0.8, + "confidence": 0.55, "canonical_issue_number": 45561, "canonical_pr_number": 45645, "issue_numbers": [ @@ -2299,9 +2426,7 @@ 45645 ], "evidence_types": [ - "closing_reference", - "shared_issue_target", - "soft_similarity" + "closing_reference" ], "pr_comparisons": [ { @@ -2612,204 +2737,209 @@ ] }, { - "cluster_id": "cluster-44001-3", - "summary": "Cluster of 3 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 44002, - "issue_numbers": [], - "pr_numbers": [ - 44001, - 44002, - 44004 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 44001, - "right_pr_number": 44002, - "code_similarity": 0.175, - "size_similarity": 0.595, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.375, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "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": [] - }, - { - "left_pr_number": 44002, - "right_pr_number": 44004, - "code_similarity": 0.125, - "size_similarity": 0.398, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.304, - "shared_filenames": [], - "shared_file_areas": [] - } - ] - }, - { - "cluster_id": "cluster-46908-2", + "cluster_id": "cluster-46970-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 46908, + "canonical_pr_number": 46986, "issue_numbers": [], "pr_numbers": [ - 46908, - 47057 + 46970, + 46986 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46908, - "right_pr_number": 47057, - "code_similarity": 0.565, - "size_similarity": 0.231, - "file_overlap": 1.0, - "area_overlap": 0.565, - "patch_similarity": 0.143, + "left_pr_number": 46970, + "right_pr_number": 46986, + "code_similarity": 0.138, + "size_similarity": 0.285, + "file_overlap": 0.194, + "area_overlap": 0.038, + "patch_similarity": 0.061, "shared_filenames": [ - "tests/models/deepseek_v32/test_modeling_deepseek_v32.py" + "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", + "src/transformers/models/granitemoehybrid/modeling_granitemoehybrid.py", + "src/transformers/models/lfm2_moe/modeling_lfm2_moe.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/qwen3_5/modeling_qwen3_5.py", + "src/transformers/models/qwen3_5_moe/modeling_qwen3_5_moe.py" ], "shared_file_areas": [ { - "filename": "tests/models/deepseek_v32/test_modeling_deepseek_v32.py", + "filename": "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", "left_ranges": [ [ - 13, - 19 + 140, + 147 + ] + ], + "right_ranges": [ + [ + 30, + 36 ], [ - 251, - 257 + 438, + 444 ], [ - 270, - 276 + 509, + 515 ], [ - 291, - 297 + 534, + 539 + ] + ] + }, + { + "filename": "src/transformers/models/granitemoehybrid/modeling_granitemoehybrid.py", + "left_ranges": [ + [ + 22, + 40 ], [ - 332, - 338 + 364, + 371 + ], + [ + 861, + 943 + ], + [ + 1063, + 1069 + ], + [ + 1074, + 1083 ] ], "right_ranges": [ [ - 13, - 19 + 35, + 41 ], [ - 248, - 260 - ], + 731, + 737 + ] + ] + }, + { + "filename": "src/transformers/models/lfm2_moe/modeling_lfm2_moe.py", + "left_ranges": [ [ - 268, - 281 + 405, + 412 + ] + ], + "right_ranges": [ + [ + 34, + 40 ], [ - 290, - 303 + 491, + 497 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 222, + 229 + ] + ], + "right_ranges": [ + [ + 38, + 44 ], [ - 332, - 345 + 586, + 592 ] ] - } - ] - } - ] - }, - { - "cluster_id": "cluster-46971-2", - "summary": "Cluster of 2 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 47058, - "issue_numbers": [], - "pr_numbers": [ - 46971, - 47058 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46971, - "right_pr_number": 47058, - "code_similarity": 0.643, - "size_similarity": 0.909, - "file_overlap": 1.0, - "area_overlap": 0.339, - "patch_similarity": 0.286, - "shared_filenames": [ - "tests/models/mamba2/test_modeling_mamba2.py" - ], - "shared_file_areas": [ + }, { - "filename": "tests/models/mamba2/test_modeling_mamba2.py", + "filename": "src/transformers/models/qwen3_5/modeling_qwen3_5.py", "left_ranges": [ [ - 418, - 424 + 412, + 417 ], [ - 450, - 456 + 429, + 436 ], [ - 480, - 497 + 825, + 831 ], [ - 508, - 514 + 1220, + 1225 ] ], "right_ranges": [ [ - 404, - 411 + 33, + 39 + ], + [ + 437, + 443 + ] + ] + }, + { + "filename": "src/transformers/models/qwen3_5_moe/modeling_qwen3_5_moe.py", + "left_ranges": [ + [ + 33, + 39 ], [ - 420, - 426 + 409, + 414 ], [ - 452, - 458 + 426, + 433 ], [ - 482, - 488 + 862, + 875 + ], + [ + 909, + 915 + ], + [ + 1292, + 1320 + ], + [ + 1328, + 1333 + ] + ], + "right_ranges": [ + [ + 33, + 39 ], [ - 503, - 509 + 434, + 440 ] ] } @@ -3093,7 +3223,16 @@ ] } ], - "duplicate_issues": [], + "duplicate_issues": [ + { + "cluster_id": "cluster-46894-4", + "canonical_issue_number": 46894, + "duplicate_issue_numbers": [ + 46897 + ], + "reason": "Issues in cluster-46894-4 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." + } + ], "duplicate_prs": [ { "cluster_id": "cluster-41211-3", @@ -3114,15 +3253,13 @@ "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-43451-4", + "cluster_id": "cluster-43451-2", "canonical_pr_number": 43451, "duplicate_pr_numbers": [ - 46707, - 46836, - 47003 + 46836 ], "target_issue_number": null, - "reason": "PRs in cluster-43451-4 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-43451-2 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-43698-3", @@ -3150,16 +3287,6 @@ "target_issue_number": 43979, "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." }, - { - "cluster_id": "cluster-44001-3", - "canonical_pr_number": 44002, - "duplicate_pr_numbers": [ - 44001, - 44004 - ], - "target_issue_number": null, - "reason": "PRs in cluster-44001-3 are treated as duplicates because they share soft_similarity evidence." - }, { "cluster_id": "cluster-44018-2", "canonical_pr_number": 44068, @@ -3185,7 +3312,7 @@ 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." + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference evidence." }, { "cluster_id": "cluster-45647-3", @@ -3216,44 +3343,47 @@ }, { "cluster_id": "cluster-46894-4", - "canonical_pr_number": 46903, + "canonical_pr_number": 46895, "duplicate_pr_numbers": [ - 46895, 46958 ], "target_issue_number": 46894, "reason": "PRs in cluster-46894-4 are treated as duplicates because they converge on issue #46894 with closing_reference, shared_issue_target, soft_similarity evidence." }, { - "cluster_id": "cluster-46908-2", - "canonical_pr_number": 46908, + "cluster_id": "cluster-46907-7", + "canonical_pr_number": 46907, "duplicate_pr_numbers": [ - 47057 + 46927, + 46949, + 46951, + 46971, + 47055, + 47056 ], "target_issue_number": null, - "reason": "PRs in cluster-46908-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46907-7 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-46926-6", - "canonical_pr_number": 46927, + "cluster_id": "cluster-46908-5", + "canonical_pr_number": 46926, "duplicate_pr_numbers": [ - 46926, - 46949, + 46908, 47054, - 47055, - 47056 + 47057, + 47058 ], "target_issue_number": null, - "reason": "PRs in cluster-46926-6 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46908-5 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-46971-2", - "canonical_pr_number": 47058, + "cluster_id": "cluster-46970-2", + "canonical_pr_number": 46986, "duplicate_pr_numbers": [ - 46971 + 46970 ], "target_issue_number": null, - "reason": "PRs in cluster-46971-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46970-2 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-47013-3", diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 003befcbb3fa29ea91da2d5d06e176a89a139a9a..05cdcd9dd4891cf70c58ca12d0fad739d169f6ac 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": "20260705T000031Z", - "generated_at": "2026-07-05T00:05:52Z", + "snapshot_id": "20260705T060758Z", + "generated_at": "2026-07-05T06:18:09Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -13,8 +13,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -132,20 +132,20 @@ "issue:45237|issue:46015", "issue:43039|issue:44805" ], - "prepared_review_unit_hash": "477c0639b9cb0093c429dedc8cf992c52264dd074dcdece1ebb5f4de9e0b8e8b", + "prepared_review_unit_hash": "82f9a1ec9fbd68097d9f773b7bb2353a92764a69633b568909e17e935b80ef31", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 63, - "serialized_chars": 46569, + "serialized_chars": 46572, "estimated_input_tokens": 11643, "estimated_eval_tokens": 23542 }, @@ -154,360 +154,330 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous. The main duplicate family is the DeepSpeed ZeRO-3 empty-tensor IndexError reports (41762, 43596, 43638), plus a separate torch 2.4.1 / transformers 5.8.0 incompatibility pair (45800/46515).", - "confidence": 0.84, - "canonical_issue_reason": "Issue 41762 is the earliest ZeRO-3 empty-tensor IndexError report in this packet, and it anchors the same failure family later seen in 43596 and 43638.", + "summary": "The cluster is mostly heterogeneous, but there are two conservative duplicate subgroups: a ZeRO-3 index-out-of-bounds/loading-failure subgroup and a docs-cleanup subgroup about v5 removals. The ZeRO-3 items look like near-duplicates, not provably exact duplicates from the packet alone.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 43638 is the clearest representative of the ZeRO-3 'index 0 is out of bounds' subgroup. It is a likely near-duplicate of 43596 and closely related to 41762, but the packet only supports a conservative same-bug grouping, not an exact-duplicate claim.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43596 has the clearest generic wording for the ZeRO-3 empty-tensor IndexError family, so it is the most reusable triage label even though it is not the earliest report.", + "best_issue_reason": "Issue 43638 is the best global issue representative because it names the core ZeRO-3 index-out-of-bounds failure most specifically and sits at the center of the strongest near-duplicate cluster in the packet.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different failures: config attribute missing vs chat/completions streaming API behavior." + "reason": "Different bugs: Qwen3.5 config attribute error vs streaming chat/completions failure." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Different model/backend problems on different hardware paths; only broad GPU/MPS similarity." + "reason": "Different model/backend problems; no shared concrete fix path." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different symptoms and subsystems: image dtype mismatch vs AMD GPU model failure." + "reason": "Both are runtime failures, but the model and failure mechanism differ." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Both are dtype/device issues, but for different models and code paths." + "reason": "Autocast dtype mismatch is not the same as the AMD GPU device-map failure." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Different models and failure modes; too broad to be duplicates." + "reason": "Different models and different failure modes; not one bug." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "AMD GPU support issue vs invalid-method bug in a different model class." + "reason": "Invalid method call in GraniteMoEHybrid and AMD GPU failure are unrelated." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch vs ONNX export error are unrelated." + "reason": "DeepSpeed collectives mismatch and ONNX export error are unrelated." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "ONNX export problem vs Qwen3MoE FP8 failure; different code paths." + "reason": "ONNX export and Qwen3Moe FP8 failure are different code paths." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "No concrete overlap beyond generic model-loading/backend failure." + "reason": "Export-to-ONNX vs auto device-mapping CUDA assert are different bugs." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Different product areas: table QA crash vs documentation example cleanup." + "reason": "Task crash and docs cleanup are not the same underlying issue." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export issue vs import-path breakage." + "reason": "ONNX export issue and GenerationMixin import error are distinct." }, { "left": "issue:44509", "right": "issue:44512", - "accept": false, - "reason": "Both are docs regressions, but they reference different removed items and likely need separate edits." + "accept": true, + "reason": "Both are docs cleanup for v5 removals, so they can be fixed together." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "Different trainer/resume problems; no shared concrete bug." + "reason": "TRL embedding unpacking bug and resume-training/trust-remote-code bug are unrelated." }, { "left": "issue:46165", "right": "issue:46869", "accept": false, - "reason": "Local-resume/trust-remote-code issue vs multi-GPU device-map issue." + "reason": "Different training/device-map problems with no shared concrete fix." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Generic mask/index shape mismatch vs Gemma4 device_map forward mismatch; not the same bug." + "reason": "Mask/index shape mismatch and Gemma4 device mismatch are different bugs." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Only loosely related IndexError wording; the actual failure messages and contexts differ." + "reason": "Both are index-ish errors, but the ZeRO-3 loading bug is a different code path from the mask-shape mismatch." }, { "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "Unrelated training resume/remote-code issue." + "reason": "ZeRO-3 loading failure and resume-from-local-model behavior are unrelated." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Both are IndexErrors, but the reported failure modes differ enough that they are not clearly the same bug." + "reason": "Same broad error family, but not the same concrete bug or code path." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Different multimodal/video loading bug; no concrete shared code path." + "reason": "Different multimodal input issues: mask indexing vs video token/feature mismatch." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Generic IndexError vs TP/MoE support issue." + "reason": "Generic index error and GPT-OSS tensor-parallel issue are unrelated." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Different features and errors; not a duplicate pair." + "reason": "Resume-training bug and broken example documentation are unrelated." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video chat-template bug vs training resume issue." + "reason": "Video StopIteration bug and training resume bug are different." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "Processor voice preset bug vs ONNX export bug." + "reason": "Processor preset bug and ONNX export error are unrelated." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both mention distributed training, but the actual errors and model contexts differ." + "reason": "Both involve DeepSpeed, but the concrete failures are different." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same DeepSpeed ZeRO-3 empty-tensor IndexError family; the titles describe the same concrete failure mode on different Bert reproductions." + "reason": "Same ZeRO-3 index-out-of-bounds loading failure; 43596 is a likely near-duplicate of 43638." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy path vs GPT-OSS tensor-parallel issue." + "reason": "Liger Kernel cross-entropy call and GPT-OSS tensor-parallel bug are unrelated." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Different issues: detokenization regression vs LoRA merge/save-load mismatch." + "reason": "Incorrect detokenization and LoRA save/load output mismatch are different issues." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Different model/runtime failures with no shared code path." + "reason": "Qwen NVFP4 runtime error and LoRA merge/save mismatch are unrelated." }, { "left": "issue:44841", "right": "issue:46661", "accept": false, - "reason": "Processor failure for a specific model vs remote-code model_class/config_class attribute error." + "reason": "Processor failure and remote-code config_class error are not the same bug." }, { "left": "issue:43716", "right": "issue:46869", "accept": false, - "reason": "Different model families and failure modes." + "reason": "Different models and backend/device-map problems." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Both involve DeepSpeed, but the underlying bugs are different." + "reason": "DeepSpeed training errors differ: weight-shape assertion vs collective mismatch." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpoint/FSDP recompute mismatch is not the same as the mask-shape IndexError." + "reason": "Activation-checkpoint recompute mismatch is not the same as a mask/indexing error." }, { "left": "issue:45685", "right": "issue:46869", "accept": false, - "reason": "Unrelated backend/hardware failures." + "reason": "MPS histogram backend issue and multi-GPU device-map issue are unrelated." }, { "left": "issue:45859", "right": "issue:45902", "accept": false, - "reason": "Same component family, but different bugs: CPU-offload incompatibility vs uninitialized-memory materialization." + "reason": "Same model family, but the CPU offload incompatibility and uninitialized-memory materialization are different bugs." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "Different model/backend problems; no shared concrete bug." - }, - { - "left": "issue:34928", - "right": "issue:36331", - "accept": false, - "reason": "Different training/runtime errors." - }, - { - "left": "issue:43596", - "right": "issue:45161", - "accept": false, - "reason": "ZeRO-3 empty-tensor IndexError vs tensor-parallel MoE issue." - }, - { - "left": "issue:43716", - "right": "issue:45507", - "accept": false, - "reason": "Different model families and failure modes." - }, - { - "left": "issue:45874", - "right": "issue:46294", - "accept": false, - "reason": "Different multimodal inference bugs." - }, - { - "left": "issue:44292", - "right": "issue:45237", - "accept": false, - "reason": "Different model/runtime failures." - }, - { - "left": "issue:39290", - "right": "issue:45081", - "accept": false, - "reason": "Different models and failures; only superficial version-era similarity." - }, - { - "left": "issue:45823", - "right": "issue:45874", - "accept": false, - "reason": "Device-map forward mismatch vs expensive reverse embedding lookup are not the same bug." - }, - { - "left": "issue:45800", - "right": "issue:46515", - "accept": true, - "reason": "Same version-compatibility problem report: torch 2.4.1 incompatibility with transformers 5.8.0/related stack." - }, - { - "left": "issue:34928", - "right": "issue:41720", - "accept": false, - "reason": "FSDP checkpointing mismatch vs Qwen3 device-mapping CUDA assert." + "reason": "ZeRO-3 Bert loading bug and GPT-OSS tensor-parallel issue are unrelated." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "General Torch compatibility issue vs Qwen NVFP4 runtime failure." + "reason": "Generic torch compatibility issue vs specific Qwen NVFP4 runtime error." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Different dtype/backend problems on different models." + "reason": "Autocast dtype mismatch and MPS integer-histogram limitation are unrelated." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Torch dynamo tracing grouped_mm failure vs image preprocessor dtype mismatch." + "reason": "Grouped_mm tracing failure and image-preprocessor dtype mismatch are different bugs." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Autocast dtype mismatch vs Qwen NVFP4 runtime failure." + "reason": "Different model/runtime failures with no shared fix." }, { "left": "issue:43716", "right": "issue:46468", "accept": false, - "reason": "Different model-specific issues." + "reason": "Different models and failure modes." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Different model/runtime errors." + "reason": "Different Qwen variants and different bugs." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "Invalid method bug vs MPS int-histogram backend limitation." + "reason": "Invalid method call in GraniteMoEHybrid and unsupported MPS op are unrelated." }, { "left": "issue:43716", "right": "issue:45685", "accept": false, - "reason": "Different model/backend failures." - }, - { - "left": "issue:43716", - "right": "issue:44292", - "accept": false, - "reason": "Different model/runtime bugs." + "reason": "Different models and backend problems." }, { "left": "issue:44918", "right": "issue:45161", "accept": false, - "reason": "Trainer embedding unpacking bug vs MoE tensor-parallel issue." + "reason": "TRL embedding unpacking bug and GPT-OSS tensor-parallel issue are unrelated." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger kernel cross-entropy issue vs trl SFT embedding unpacking issue." + "reason": "Liger Kernel cross-entropy issue and TRL SFT trainer bug are unrelated." }, { "left": "issue:41762", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 load-time empty-tensor IndexError family; Gemma3 and Bert are different repros of the same underlying bug." + "reason": "Same ZeRO-3 'index 0 out of bounds' loading failure, just reported on a different model." }, { "left": "issue:43827", "right": "issue:44512", + "accept": true, + "reason": "Both are docs references left behind after v5 removals." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": true, + "reason": "Both describe the same torch 2.4.1 incompatibility." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Both are Gemma4 bugs, but the device mismatch and reverse-embedding lookup are different problems." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Different models and failures: Gemma3 config attribute vs Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Repeated pair is still unrelated: autocast mismatch vs unsupported MPS op." + }, + { + "left": "issue:43039", + "right": "issue:44805", "accept": false, - "reason": "Both are docs regressions, but they concern different removed APIs/commands." + "reason": "Liger Kernel cross_entropy issue and mask/index shape error are not the same bug." } ] }, "evaluator_result": { "accept": false, - "feedback": "The soft-edge acceptances look conservative, but the canonical_issue_reason is not fully grounded: issue 43596 is not the earliest ZeRO-3 empty-tensor IndexError in the packet (issue 41762 predates it). Please remove or correct that chronology claim before accepting." + "feedback": "The overall clustering is plausible, but the writeup overstates the evidence in a few places. In particular, calling issue 43596 an \"exact match\" to 43638 and describing an \"exact duplicate relationship\" is stronger than what the packet supports: the titles are highly similar and the issues look like likely duplicates, but there is no explicit duplicate linkage or body text here proving exact sameness. Please soften that language to \"near-duplicate\" / \"likely duplicate\" and keep the subgroup claim conservative." }, "evaluator_used": true, "retried": true, @@ -523,8 +493,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -621,292 +591,292 @@ "issue:43572|issue:45070", "issue:45800|issue:45823" ], - "prepared_review_unit_hash": "cb8a4971ae2a0513c1e26fb652cdd2e666de8cf526e03037bf33bb46f7d6b243", + "prepared_review_unit_hash": "68c40c7a65f2cf3fcaf76bdc02d931deb39307bbd2bbd871b1dc2910f580424b", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 42, - "serialized_chars": 40047, - "estimated_input_tokens": 10012, - "estimated_eval_tokens": 20280 + "serialized_chars": 40042, + "estimated_input_tokens": 10011, + "estimated_eval_tokens": 20278 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of unrelated issues. The only clear duplicate families are the DeepSpeed ZeRO-3/BERT load failures and the v5 AutoTokenizer token-ID regression; the rest are only loosely similar by subsystem or error style.", - "confidence": 0.85, - "canonical_issue_reason": "issue:45920 is the strongest canonical issue: it has the broadest and clearest title, describes a concrete v5 tokenizer regression across multiple models, and is more representative than the narrower Granite-only report.", + "summary": "This set is mostly heterogeneous issue noise. I found two likely duplicate clusters: the DeepSpeed ZeRO-3 load/indexing failures and the v5 AutoTokenizer wrong-token-id regression. All other soft pairs look unrelated.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45920 is the broadest and clearest representative of the tokenizer regression cluster; it covers multiple model families and is more general than issue:45812.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best global issue candidate because it is broad, actionable, and clearly framed as a regression affecting multiple model families, making it a good anchor for the duplicate family.", + "best_issue_reason": "issue:45920 is the best global issue candidate here because it most clearly captures a recurring, cross-model v5 tokenizer bug that the paired report also points to.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Both mention torch/transformers breakage, but they describe different incompatibilities and likely different code paths." + "reason": "Both mention torch/version breakage, but they describe different incompatibilities and different failure modes." }, { "left": "issue:43866", "right": "issue:46402", "accept": false, - "reason": "One is a corrupted checkpoint report; the other is an unmapped text-encoder weight loading bug. Not the same failure." + "reason": "Different models and different bugs: corrupted checkpoint vs missing text-encoder weights." }, { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "Different models and different symptoms: vLLM/Gemma3 loading versus distributed build/runtime support in v5.9.0." + "reason": "Different stack and symptom; vLLM/PaddleOCR-VL load failure is unrelated to the USE_DISTRIBUTED=0 PyTorch support break." }, { "left": "issue:41762", "right": "issue:43596", "accept": true, - "reason": "Both report the same ZeRO-3 init/load-time IndexError class and appear to stem from the same loading bug." + "reason": "Both report the same ZeRO-3 initialization/load path producing an index-out-of-bounds on empty tensors during model loading." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Gemma3 sliding-window config missing and Qwen3 auto-device-map cuda assert are unrelated bugs." + "reason": "Gemma3 sliding_window_pattern missing in vLLM and Qwen3 auto-device-map cuda assert are different code paths and failures." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs issues, but they concern different API/documentation mismatches and do not share the same underlying bug." + "reason": "Both are docs-related, but they cover different outdated references and are not the same documentation defect." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to the torch USE_DISTRIBUTED=0 compatibility problem." + "reason": "Unrelated runtime errors: missing Float8 storage vs PyTorch build/distributed support break." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "AMD GPU support failure and table-question-answering crash are unrelated." + "reason": "Different models/features and different failures; no evidence of a shared underlying bug." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger kernel cross-entropy fallback and Qwen3-VL video StopIteration are different code paths." + "reason": "Liger Kernel cross_entropy dispatch and Qwen3-vl video StopIteration are unrelated." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Incorrect detokenization and outdated RAG docs are unrelated." + "reason": "A detokenization regression is unrelated to outdated RAG examples." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Device-mapping forward mismatch and video token/template mismatch are different multimodal bugs." + "reason": "Device-map forward mismatch and chat-template video token mismatch are different issues." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Processor failure for Voxtral and Gemma4 PLE device mismatch are unrelated." + "reason": "Processor failure for Voxtral and Gemma4 device mismatch are distinct bugs." }, { "left": "issue:43866", "right": "issue:46772", "accept": false, - "reason": "Corrupted checkpoint and ModelOpt compatibility are not the same issue." + "reason": "Corrupted checkpoint and NVFP4 compatibility are different problems." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "Checkpoint loading failure and RAG example docs are unrelated." + "reason": "NemotronH checkpoint loading failure is unrelated to outdated RAG examples." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Remote-code regressions and USE_DISTRIBUTED=0 support are different compatibility problems." + "reason": "Both are regressions in recent versions, but they affect different subsystems and error modes." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Both describe the same v5 AutoTokenizer wrong-token-ID regression; 45920 is the broader multi-model instance." + "reason": "Both describe the same v5 tokenizer regression causing wrong token IDs across multiple model families; 45920 is the broader report." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "A corrupted checkpoint is not the same as a model implementation load failure." + "reason": "Both are model-loading failures, but they concern different models and likely different root causes." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "Qwen NVFP4 runtime error and RAG docs are unrelated." + "reason": "Qwen-3 NVFP4 runtime failure is unrelated to RAG documentation drift." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes Int8Params incompatibility is unrelated to the pytorch/distributed build issue." + "reason": "bitsandbytes Int8Params API break is unrelated to the PyTorch distributed build issue." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding_window, but they affect different models and different config bugs." + "reason": "Both mention sliding_window, but the models, stack, and failures are different enough that this does not look like one bug." }, { "left": "issue:41093", "right": "issue:43638", - "accept": true, - "reason": "Both are the same ZeRO-3/BERT-style IndexError during loading/training and look like the same underlying bug." + "accept": false, + "reason": "Different error shapes and different contexts; one is a mask-length mismatch, the other a zero-3/empty tensor index error." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper loading/runtime failure and table-question-answering crash are unrelated." + "reason": "Whisper runtime failure and table-question-answering crash are unrelated." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "Unit-test loading failure and corrupted checkpoint are different issues." + "reason": "Different model-specific loading problems with no clear shared root cause." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "General torch incompatibility and pydantic PretrainedConfig breakage are not the same bug." + "reason": "Torch 2.10 compatibility and pydantic PretrainedConfig breakage are separate compatibility issues." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Detokenization regression and table-question-answering crash are unrelated." + "reason": "Detokenization regression and table QA crash are unrelated." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper loading failure and pydantic model-field breakage are unrelated." + "reason": "Whisper load/runtime failure does not match the pydantic config regression." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Qwen3.5 chat crash and streaming chat/completions API failure are different symptoms and likely different paths." + "reason": "Both involve Qwen3.5 chat behavior, but one is transformers chat and the other is streaming chat/completions; too different to merge." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "ZeRO-3 IndexError and torch/transformers version incompatibility are unrelated." + "reason": "ZeRO-3 load/index error is unrelated to the torch/transformers version incompatibility." }, { "left": "issue:43541", "right": "issue:46869", "accept": false, - "reason": "Mixtral grouped_mm tracing failure and YOLO multi-GPU device-map failure are different issues." + "reason": "Mixtral torch-dynamo grouped_mm tracing failure is unrelated to YOLO multi-GPU device-map auto failure." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "NVFP4 runtime error and table task crash are unrelated." + "reason": "Qwen NVFP4 runtime issue and table QA crash are not the same bug." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Gemma3 sliding-window config and Gemma4 device-mapping are unrelated." + "reason": "Different models and different failure modes: missing config attr in vLLM versus Gemma4 forward device mismatch." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Streaming inference API failure and MPS histogram Int error are unrelated." + "reason": "Qwen3.5 streaming API failure is unrelated to the MPS histogram_int error." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "ImportError for GenerationMixin and Gemma3 sliding-window config are unrelated." + "reason": "ImportError for GenerationMixin and a Gemma3 sliding_window config issue are unrelated." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Both involve MoE-related models, but the traced grouped_mm failure and MPS histogram failure are different code paths." + "reason": "Mixtral grouped_mm tracing and MPS histogram_int errors are unrelated." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Mixtral tracing error and autocast dtype mismatch are different bugs." + "reason": "Different MoE-related symptoms, but different models, execution modes, and likely root causes." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure for Voxtral and template compilation error are unrelated." + "reason": "Processor loading failure and template-node compilation error are unrelated." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "lm_head weight-tying bug and corrupted checkpoint are not the same failure." + "reason": "Tied lm_head weights in Mistral3 is unrelated to a corrupted Ovis2 checkpoint." }, { "left": "issue:45507", "right": "issue:46869", "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid and YOLO multi-GPU auto device-map failure are unrelated." + "reason": "GraniteMoEHybrid invalid-method calls and YOLO multi-GPU device-map failure are different bugs." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are DeepSpeed ZeRO-3 errors, but the specific exceptions and model contexts differ enough that they do not look like one bug." + "reason": "Mask-shape mismatch and ZeRO-3 zero-sized index error are different failure signatures." }, { "left": "issue:45399", "right": "issue:46661", "accept": false, - "reason": "Flash-attn fallback gating and remote-code config_class failures are different issues." + "reason": "Flash-attn fallback gating and remote-code config_class AttributeError are unrelated." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and a broken PretrainedConfig field in pydantic are related only by version churn, not the same bug." + "reason": "Missing pad_token_idx in StableLmConfig is a different config issue than the pydantic PretrainedConfig regression." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch/transformers compatibility and Gemma4 device mismatch are unrelated." + "reason": "Torch/transformers compatibility break is unrelated to Gemma4 device placement mismatch." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: it picks 45920 as the broadest tokenizer-regression issue and correctly characterizes most other pairs as unrelated by title/context. The accepted soft pairs are conservatively phrased as likely same-family DeepSpeed/ZeRO-3 or tokenizer regressions, without overclaiming beyond the evidence provided." + "feedback": "The summary is well grounded in the packet and stays conservative. The two accepted duplicate clusters are supported by the titles: ZeRO-3 load/index-out-of-bounds (41762/43596) and the v5 AutoTokenizer wrong-token-id regression (45812/45920). The remaining soft-edge rejects are generally defensible as different models, stacks, or failure modes." }, "evaluator_used": true, "retried": false, @@ -922,8 +892,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -1008,14 +978,14 @@ "issue:45183|issue:46463", "issue:44918|issue:45507" ], - "prepared_review_unit_hash": "dbd75fd4ef2cbfe34871e5c79ba6eed7ac80e03be5cf7add41068fc8ef748eba", + "prepared_review_unit_hash": "9fbe597630892ba93fb13dd2ca5fca6a3fd4f54febb702235f99fa86dd1ff67d", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, @@ -1030,208 +1000,205 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are unrelated bug reports with similar wording. The only clear duplicate-style match is the trust_remote_code/custom-module cache collision pair (45632/45698).", - "confidence": 0.88, - "canonical_issue_reason": "issue:45632 is the cleaner root-cause report: it describes the cache-path collision for local models with the same leaf directory name, which also explains the wrong-module symptom in issue:45698.", + "summary": "This cluster is mostly a collection of distinct Transformers bug reports spanning model loading, multimodal preprocessing, quantization, cache behavior, and validation errors. The soft pairs share broad themes but do not look like the same underlying defect, so none are accepted.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45020 is the broadest, clearest regression report in the set: recent versions breaking remote_code models. It is concrete, user-facing, and representative of the cluster\u2019s recurring loading regressions.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45632 is the best representative because it states the underlying bug more generally and would absorb the downstream manifestation reported in issue:45698.", + "best_issue_reason": "issue:45020 is the most generally useful canonical issue because it captures a high-impact, cross-model failure mode with clear actionable scope.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different failures: PaddleOCR-VL loading under transformers/vLLM vs GGUF gpt-oss architecture support. Not the same code-path." + "reason": "Different bugs: PaddleOCR-VL model load failure versus gpt-oss GGUF architecture support." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Both mention shape/indexing errors, but one is a mask-length mismatch and the other is Qwen3.5 embedding unpacking with TRL SFT trainer; different bugs." + "reason": "Both are shape-related failures, but one is a mask/tensor index mismatch and the other is a TRL SFT unpacking issue." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Grouped_mm tracing error in Mixtral vs tensor-parallel not working for GPT-OSS MoE; different execution paths and symptoms." + "reason": "Different code paths: torch dynamo/grouped_mm tracing failure versus tensor-parallel-only GPT-OSS MoE behavior." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Both are dtype-related, but one is autocast on Phi-tiny-MoE and the other is chat/completions streaming on Qwen3.5-0.8B. Not the same bug." + "reason": "Unrelated symptoms: autocast dtype mismatch in Phi-tiny-MoE versus streaming chat/completions failure in Qwen3.5." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Memory/OOM under int4 quantization is unrelated to tied-weights issues on translation models." + "reason": "One is an int4 quantization memory regression; the other is a tied-weights issue on translation models." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3.5 streaming inference are distinct model/runtime problems." + "reason": "GGUF architecture support is unrelated to Qwen3.5 streaming inference in chat/completions." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "`init_empty_weights` argument mismatch is a loader/init bug; pydantic `PretrainedConfig` field breakage is a separate API/serialization issue." + "reason": "Different regressions: unexpected _is_hf_initialized arg in init_empty_weights versus pydantic PretrainedConfig field breakage." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to the pydantic `PretrainedConfig` breakage." + "reason": "Qwen3-MoE sliding_window behavior is not the same bug as the v5.4.0 pydantic model break." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are IndexErrors, but one is a mask/tensor length mismatch and the other is a zero-sized tensor during deepspeed zero3 init. Different causes." + "reason": "Both are IndexErrors, but the failing shapes and initialization contexts are different." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Image preprocessor dtype mismatch in Mistral-3 vs video token/feature mismatch in Qwen3.5-9B; different modalities and code paths." + "reason": "Both involve multimodal models, but one is an image-preprocessor dtype mismatch and the other is a video feature/token count mismatch." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote-code model breakage and torch/transformers incompatibility are too broad and unrelated to be the same concrete bug." + "reason": "Remote-code compatibility breakage is not the same as the torch 2.4.1 / transformers 5.8.0 incompatibility report." }, { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Siglip2 text embedding quality regression and nonsensical AutoModel/pipeline output are not the same underlying issue." + "reason": "Same model family, but one reports degraded text embeddings and the other nonsensical AutoModel/pipeline outputs; not clearly the same defect." }, { "left": "issue:44479", "right": "issue:46294", "accept": false, - "reason": "Video-input regression in v5.3.0 affects a family of Qwen models, but it is not the same as the apply_chat_template video-feature/token mismatch." + "reason": "Both are video-related, but the regression scope and failure modes differ." }, { "left": "issue:46468", "right": "issue:46869", "accept": false, - "reason": "Nemotron example breakage and YOLO multi-GPU device_map_auto failure are unrelated model-integration issues." + "reason": "Official Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Qwen multimodal `text_config` unwrapping and Qwen3_5MoeTextRotaryEmbedding uninitialized memory are different bugs." + "reason": "Both touch Qwen multimodal internals, but one is text_config unwrapping and the other is uninitialized inv_freq after to_empty materialization." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "Deepspeed zero3 init index error is unrelated to resuming training with local models and trust_remote_code." + "reason": "Zero3/BERT indexing failure is unrelated to resuming training with locally downloaded remote_code models." }, { "left": "issue:46640", "right": "issue:46661", "accept": false, - "reason": "Flash Attention 2 integration problem and remote-code model_class/config_class AttributeError are different failures." + "reason": "Flash Attention 2 fallback issues and remote-code model_class/config_class errors are different code paths." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "One is an image dtype mismatch; the other is TRL unpacking failure for Qwen3.5 inputs. Not the same bug." + "reason": "Different failures: multimodal dtype mismatch versus TRL SFT embedding unpacking." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP weight-shape runtime error and Qwen3.5 embedding unpacking failure are unrelated." + "reason": "FSDP weight dimensionality error is unrelated to the TRL SFT trainer embedding issue." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Jupyter custom-model initialization in Transformers 5.0.0 is not the same as flash-attn2 fallback gating." + "reason": "Jupyter custom-model initialization and FlashAttention fallback blocking are distinct regressions." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-quality reports with no shared code-path." + "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-specific issues." }, { "left": "issue:45632", "right": "issue:45698", - "accept": true, - "reason": "Both describe the same trust_remote_code/custom-module cache collision leading to the wrong custom module being loaded; one is the root-cause report and the other is the manifested wrong-module load." + "accept": false, + "reason": "Both involve remote_code/custom-module loading, but one is a leaf-directory cache collision while the other is wrong module loading after save_pretrained; not enough to merge." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Mistral3 tied lm_head weights and translation-model tied-weights issues are not the same concrete bug." + "reason": "Mistral3 lm_head tie-breaking affects a different model and path than the translation-model tied-weights regression." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "FineGrainedFP8Config failure on Qwen3Moe is unrelated to flash-attn2 fallback checks." + "reason": "FineGrainedFP8Config failure and FlashAttention fallback blocking are unrelated." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to tied-weight handling on translation models." + "reason": "Quantization/offloading behavior is a different issue from tied weights on translation models." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A DSA indexer discussion and a docs typo are not duplicates." + "reason": "A DSA indexer activation bug and a docs typo are not the same underlying change." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache/linear_attention crash on Qwen3.5 is unrelated to tied weights in translation models." + "reason": "PagedAttentionCache linear_attention crash on Qwen3.5 is unrelated to tied-weight handling in translation models." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Cache-strategy typo and docs mention of removed run command are different documentation issues." + "reason": "A cache-strategy typo and a removed run-command doc reference are separate documentation issues." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading failure is unrelated to tied-weight issues." + "reason": "NemotronH checkpoint loading failure is unrelated to tied weights in translation models." }, { "left": "issue:45183", "right": "issue:46463", "accept": false, - "reason": "Generic input_ids/inputs_embeds XOR validation messaging is broader than the T5 decoder-input validation bug; different scopes and fixes." + "reason": "The generic input_ids/inputs_embeds XOR validation issue and T5 decoder-input validation are related in spirit but not the same bug." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Qwen3.5 embedding unpacking with TRL SFT trainer is unrelated to GraniteMoEHybrid calling an invalid method." + "reason": "Qwen3.5 TRL embedding unpacking and GraniteMoEHybrid invalid method calls are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted pair, issue:45632/issue:45698, is well supported by the titles and the stated cache-path/custom-module relationship, while the other soft-edge rejections do not overclaim from the available 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 }, @@ -1243,8 +1210,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -1328,20 +1295,20 @@ "issue:43824|issue:43872", "issue:43782|issue:44960" ], - "prepared_review_unit_hash": "23840281e09dd6df45095fcf78b273b25c325fcca411331aca03b4087e25d95b", + "prepared_review_unit_hash": "6512e6351e1c37c8217add90b99bb2362ffa3fd98d60f5214ed122826cd52fe4", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36285, + "serialized_chars": 36287, "estimated_input_tokens": 9072, "estimated_eval_tokens": 18400 }, @@ -1350,207 +1317,201 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many separate Transformers regressions around loading, tokenization, remote-code, quantization, and model-specific errors. Only a few soft edges look like true duplicates, mainly where the titles point to the exact same underlying contract/regression.", - "confidence": 0.73, - "canonical_issue_reason": "Issue #45020 is the broadest umbrella for the cluster\u2019s repeated loading/tokenizer/remote-code regression reports, so it is the best single issue to anchor triage even though the cluster is not a tight duplicate set.", + "summary": "The cluster is still mostly a mix of unrelated bug reports, but the non-persistent-buffer regression is a small three-way duplicate cluster: #43644, #43950, and #44534 all describe the same Transformers v5 corruption of `register_buffer(persistent=False)` state. The other soft-edge candidates remain too broad or too code-path-specific to merge.", + "confidence": 0.89, + "canonical_issue_reason": "Issue #43950 is the best canonical issue because it states the regression at the right level of specificity (`from_pretrained()` corrupts non-persistent buffers in Transformers 5.x) while cleanly subsuming the near-duplicate reports #43644 and #44534.", "canonical_pr_reason": null, - "best_issue_reason": "#45020 is the most representative high-level issue in this set: it captures the recurring recent-version loading breakages that many nearby issues are variants of, making it the best global issue anchor.", + "best_issue_reason": "Issue #43950 is the strongest representative of the duplicate buffer-corruption cluster: it is the most explicit about the `from_pretrained()` failure mode and general enough to cover the paraphrased reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Different failures: DeepSeek-Coder detokenization vs Qwen3 NVFP4 load/runtime error." + "reason": "Both are loading/tokenization failures, but they involve different models and likely different code paths; not the same bug." }, { "left": "issue:46489", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they describe different concrete bugs and model behaviors; not the same underlying fix." + "reason": "Both are tokenizer regressions, but they affect different models and different failure modes; too different to merge." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Doc example/model-processor mismatch and quantization failure vs SigLIP2 text embedding quality regression." + "reason": "Same model family only in a broad sense; one is a docs/example mismatch with quantization failure, the other is a text-embedding quality regression." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Same non-persistent buffer corruption regression in Transformers v5; 44534 is essentially the same bug restated." + "reason": "These are the same non-persistent-buffer regression in Transformers v5, described with slightly different wording." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Unrelated trainer API change vs DeepSpeed ZeRO-3 Gemma3 loading error." + "reason": "A trainer API compatibility error and a DeepSpeed ZeRO-3 loading index error are unrelated failures." }, { - "left": "issue:43644", + "left": "issue:43643", "right": "issue:43950", "accept": false, - "reason": "Different subsystems: remote-code AutoConfig field loss vs non-persistent buffer corruption." + "reason": "One is missing fields from `AutoConfig.from_pretrained(..., trust_remote_code=True)`; the other is buffer corruption during `from_pretrained()`. Different problems." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Different bitsandbytes/float8 incompatibility errors with different root causes." + "reason": "One is a bitsandbytes parameter signature incompatibility; the other is a missing storage class error. Not the same bug." }, { "left": "issue:46899", "right": "issue:46932", - "accept": true, - "reason": "Same Gemma4 quantized vision/audio embedder dtype-casting bug; 46932 explicitly references the same issue family." + "accept": false, + "reason": "These are closely related Gemma4 quantization reports, but #46932 reads like a follow-up/fix note rather than a clearly separate duplicate of the exact same defect." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Different model/pathology: autocast dtype mismatch on Phi MoE vs Qwen3.5 3D position-id/attention issue." + "reason": "Different symptoms and subsystems: dtype mismatch with autocast vs a flash-attn crash from 3D `position_ids` handling." }, { "left": "issue:45920", "right": "issue:46489", "accept": false, - "reason": "Related tokenizer regressions, but not the same concrete bug or fix scope." + "reason": "Both involve tokenizer regressions, but they affect different model families and different failure modes." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding quality issue vs broad tokenizer-version change report; not the same bug." + "reason": "Both are about tokenization/behavior changes, but one is a SigLIP2 embedding-quality issue and the other is a broad version-dependent tokenization change." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Different SigLIP2 symptoms: unexpected zero probabilities vs worse text embeddings." + "reason": "Related SigLIP/SigLIP2 surface area, but one is zero probabilities and the other is worse text embeddings; not the same concrete bug." }, { "left": "issue:45987", "right": "issue:46858", - "accept": true, - "reason": "Same StaticCache.get_seq_length contract problem returning a tensor instead of a plain int, causing downstream breakage." + "accept": false, + "reason": "Both concern `get_seq_length()`, but one is an API contract mismatch and the other is backend-specific tensor arithmetic breakage." }, { "left": "issue:43950", "right": "issue:45902", "accept": false, - "reason": "Both involve non-persistent buffers, but the second is about meta->to_empty materialization and uninitialized rotary state, which is a different bug." + "reason": "Both involve non-persistent buffers, but #45902 is a specific `inv_freq` uninitialized-memory case after `to_empty(cuda)`, not the same regression as #43950." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Both mention Qwen3.5, but one is a flash-attn crash from 3D position_ids and the other is CPU-offload incompatibility in rotary embedding." + "reason": "One is a flash-attention crash due to 3D `position_ids`; the other is CPU offload incompatibility in rotary embedding. Different code paths." }, { "left": "issue:43054", "right": "issue:43493", "accept": false, - "reason": "Different SigLIP2 complaints: degraded text embeddings vs HF/JAX implementation discrepancy." + "reason": "Both are SigLIP2-related, but one reports worse embeddings and the other compares HF vs JAX implementation discrepancy; not enough to merge." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "Different training problems: LoRA gradient checkpointing vs int4 quantization memory growth/OOM." + "reason": "Both mention memory, but one is gradient checkpointing with PEFT LoRA and the other is int4 quantization reserved-memory growth. Different defects." }, { "left": "issue:45663", "right": "issue:45902", "accept": false, - "reason": "Different Gemma4/Qwen3 issues with unrelated code paths and symptoms." + "reason": "Gemma4 FSDP2 shared-state rebuilding and Qwen3.5 non-persistent buffer initialization are unrelated." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Bad AutoTokenizer error message for Voxtral vs Molmo attribute error; unrelated." + "reason": "A bad tokenizer import error for Voxtral and a missing tied-weights attribute for Molmo are different issues." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise vs a specific model failing to load; not the same issue." + "reason": "One is generic load-time noise; the other is inability to load a specific model. Not the same underlying bug." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Tokenizer detokenization bug vs Python 3.13 flash_attn import error." + "reason": "Different models and different failure types: detokenization regression vs `import_utils` KeyError on Python 3.13." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "DeepSpeed training weight-shape error vs early-stopping metric selection issue." + "reason": "Training DeepSpeed weight-shape runtime error and early-stopping metric misconfiguration are unrelated." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Different model-loading reports with different root causes." + "reason": "Both are loading-related, but one is noisy logging and the other is a model loading failure for a specific checkpoint." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Both touch remote-code loading, but one is a specific missing-fields AutoConfig bug and the other is a broad umbrella report; not clearly the same concrete defect." + "reason": "Both mention remote code / config loading regressions, but one is missing fields in `AutoConfig` return objects while the other is a broader compatibility report." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Different memory/training issues: LoRA checkpointing inefficiency vs Qwen3 MoE memory usage." + "reason": "Both are memory-efficiency concerns, but one is PEFT/gradient checkpointing and the other is Qwen3 MoE training memory usage." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Different model/loading failures: Gemma3 ZeRO-3 index error vs Qwen3 MoE FP8 config failure." + "reason": "One is a ZeRO-3 load-time index error for Gemma3; the other is a FineGrainedFP8Config failure for Qwen3 MoE." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression vs flash-attn2 fallback gating; unrelated." + "reason": "Tokenizer load failure for a specific model and flash-attn fallback logic are unrelated." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Different model-specific loading errors with no shared bug." + "reason": "Both are model-loading issues, but one is an AutoTokenizer import error and the other is a `glm-4v-9b` loading failure; no shared root cause is evident." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Import error for a missing class vs bitsandbytes keyword-argument incompatibility." + "reason": "An unavailable model class import and a bitsandbytes parameter incompatibility are different failures." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight-only load error vs unrelated GLM5 issue." - }, - { - "left": "issue:44910", - "right": "issue:44928", - "accept": true, - "reason": "Same underlying Qwen3.5 3D position_ids handling bug in attention; one manifests as flash-attn crash, the other as SDPA fallback/NaN instability." + "reason": "Qwen3VL weight-only loading error and a generic GLM5 issue are not the same bug." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately conservative about the cluster being heterogeneous. The accepted soft edges are all plausible same-bug or same-regression families from the titles, and the rejected ones avoid overclaiming where the evidence only shows related symptoms rather than a shared fix." + "accept": false, + "feedback": "The summary is mostly conservative, but it overstates the duplicate structure. The packet also contains issue 43644 ('Transformers 5.0.0 fills non-persistent buffers with junk'), which is essentially the same non-persistent-buffer regression family as 43950/44534 and is not accounted for in the canonicalization. Because of that, the claim that there is only one clear duplicate pair and that #43950 is the best/general representative is not fully grounded in the packet. The rest of the soft-edge rejections look reasonably conservative." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1563,8 +1524,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -1647,211 +1608,211 @@ "issue:41093|issue:44805", "issue:24643|issue:30064" ], - "prepared_review_unit_hash": "d7fb8bfb06ae1482bfecc639a971bb33926295a634a898770c7f934f17109dd1", + "prepared_review_unit_hash": "c6e1effece72fdd07ee8ec6048f30cca98be4ea18dda002bf6a33280980c6fef", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36747, - "estimated_input_tokens": 9187, - "estimated_eval_tokens": 18630 + "serialized_chars": 36752, + "estimated_input_tokens": 9188, + "estimated_eval_tokens": 18632 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a mixed cluster with only a few true duplicate-style pairs: docs cleanup around removed pipeline tasks, the `_is_hf_initialized`/`Int8Params` loading bug, and the repeated mask-shape mismatch issue. Most other pairs are only thematically related (loading, tokenizer, DeepSpeed, or model-family regressions) but do not look like the same concrete bug or mergeable fix.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45020 is the broadest representative of the largest recurring theme in the cluster (recent-version remote_code/loading regressions), but the cluster is heterogeneous, so it should be treated as a loose anchor rather than a universal canonical.", + "summary": "Most pairs are false positives across unrelated subsystems, but a few are genuine duplicates: docs cleanup after v5 pipeline removals, the `_is_hf_initialized` init-empty-weights regression, Qwen3 ZeRO-3 weight-loading failures, and the training-loss double-shift bug.", + "confidence": 0.76, + "canonical_issue_reason": "issue:46894 is the clearest umbrella report for the training-loss double-shift regression and is the better representative of that duplicate pair than the narrower Florence2 report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best single issue representative because it is the most umbrella-like report, even though several nearby items are separate bugs rather than duplicates.", + "best_issue_reason": "issue:46894 has the broadest, most reusable description of the underlying bug pattern and the strongest case for serving as the cluster\u2019s canonical issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "Different failures: MPS int histogram support vs device_map='auto' forward device mismatch." + "reason": "Different bugs: MPS histogram/int issue vs device_map forward mismatch." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer backend crash and flash-attn fallback gating are different code paths." + "reason": "Different code paths: tokenizer regex crash vs flash-attn fallback gating." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Remote_code loading regression vs Gemma4 PLE device mismatch; not the same bug." + "reason": "Broad remote_code regression vs Gemma4 PLE device placement bug." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer-mapping edge case vs generic remote_code loading regression; different underlying failures." + "reason": "Related area, but not the same concrete failure; add-new-model-like tokenizer mapping bug vs generic remote_code breakage." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Autocast dtype mismatch and device placement mismatch are distinct problems." + "reason": "Different models and symptoms: autocast dtype mismatch vs device placement mismatch." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "Branch-compatibility/LoRA loading issue vs a tie_word_embeddings warning; not a duplicate." + "reason": "LoRA/Qwen3_moe compatibility issue vs tied-embeddings warning; not the same bug." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Different models and different missing-attribute failures during loading." + "reason": "Both model-loading failures, but different underlying problems and model classes." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Remote_code regression vs Kimi tokenizer codec/regression issue; separate fixes." + "reason": "Generic remote_code breakage vs Kimi tokenizer codec/regex regression." }, { "left": "issue:45715", "right": "issue:46315", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different APIs and performance paths." + "reason": "Both tokenizer performance regressions, but different APIs and likely different hot paths." }, { "left": "issue:43872", "right": "issue:46869", "accept": false, - "reason": "Int8Params/_is_hf_initialized TypeError vs YOLO multi-GPU device_map failure; unrelated." + "reason": "Unrelated: bitsandbytes/_is_hf_initialized incompatibility vs YOLO multi-GPU device map failure." }, { "left": "issue:45759", "right": "issue:46661", "accept": false, - "reason": "Composite-config unwrapping bug vs missing config_class on remote-code models; different issues." + "reason": "Different remote-code loader bugs: text_config unwrapping vs missing config_class." }, { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "Same docs regression: v5 removed these pipeline tasks, but the docs still reference them." + "reason": "Same docs regression: both report removed v5 pipeline tasks (summarization/translation/text2text-generation) still referenced in documentation." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Same `_is_hf_initialized` / Int8Params loading bug when using init_empty_weights." + "reason": "Same `_is_hf_initialized` TypeError during model loading/init_empty_weights; 44291 is the generalized report of the same regression." }, { "left": "issue:45488", "right": "issue:46489", "accept": false, - "reason": "Pattern-similar tokenizer regressions, but not clearly the same concrete DeepSeek bug or fix." + "reason": "Same broad tokenizer area, but not enough evidence they are the same underlying bug or fix target." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Tied lm_head weight bug vs non-persistent buffer corruption; different state-handling failures." + "reason": "Different problems: lm_head tying bug vs silent corruption of non-persistent buffers." }, { "left": "issue:46894", "right": "issue:46897", - "accept": false, - "reason": "Similar training-loss double-shift wording, but the evidence does not show a single shared bugfix." + "accept": true, + "reason": "Same training-loss double-shift bug pattern; Florence2 looks like a model-specific instance of the same regression." }, { "left": "issue:45020", "right": "issue:46661", "accept": false, - "reason": "Both are remote-code related, but one is a loading regression and the other is a config_class attribute failure." + "reason": "Not the same concrete bug; broad remote_code regressions should not be merged with the specific config_class AttributeError report." }, { "left": "issue:43792", "right": "issue:44220", "accept": false, - "reason": "Whisper runtime failure vs `_torch_extract_fbank_features()` issue; different audio code paths." + "reason": "Both audio/Whisper-related, but the titles point to different failures and code paths." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "One is missing weight conversion, the other is gradient averaging under GAS; not the same bug." + "reason": "Different issues: Qwen3 MOE weight conversion/loading vs gradient averaging under DeepSpeed GAS." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Gradient-checkpointing ineffective vs missing supports_gradient_checkpointing flag; related area, different defect." + "reason": "LoRA gradient-checkpointing effectiveness vs missing supports_gradient_checkpointing flag; related theme, different bug." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Trainer signature breakage vs tensor-mask shape mismatch; unrelated failures." + "reason": "Both mention shape errors, but the model/training contexts and exact mismatches are different." }, { "left": "issue:43940", "right": "issue:45313", - "accept": false, - "reason": "Both are DeepSpeed weight-loading complaints, but the concrete missing-weight causes differ." + "accept": true, + "reason": "Same DeepSpeed ZeRO-3 weight-loading failure pattern across Qwen3 family variants; likely one loader regression." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model-loading failures for different architectures/checkpoints." + "reason": "Different model-loading failures for different models; too little evidence of a shared root cause." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Bad error message vs hard tokenizer load failure; not the same bug." + "reason": "Different tokenizer failures: bad Voxtral error message vs est-roberta tokenizer loading breakage." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "RLHF NaN / SDPA fallback problem vs flash_attention crash on s_aux=None; distinct failures." + "reason": "Unrelated: RLHF gradient explosion/SDPA fallback vs flash_attention s_aux=None crash." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both are video-related, but one is a broad regression and the other is a specific StopIteration in Qwen3-vl-embedding." + "reason": "Both are video regressions, but the affected models and reported failures differ enough to avoid merging." }, { "left": "issue:41093", "right": "issue:44805", - "accept": true, - "reason": "Same mask/tensor shape-mismatch indexing failure, just with different reported sizes." + "accept": false, + "reason": "Same generic error shape, but different reports and likely different underlying causes." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "Completely different topics: DeepSpeed training weight shape error vs void segmentation-map processing." + "reason": "Completely different subsystems: DeepSpeed training error vs image processor segmentation-map failure." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall cluster summary is broadly grounded, but I\u2019d reject because a few soft-edge accepts are too strong for the evidence in the packet. 43827/44509 and 43872/44291 are solid duplicate-style matches, but 45488/46489, 43940/45313, and 46894/46897 look only pattern-similar from the titles, not clearly the same bugfix or single PR-worthy issue. Given the packet\u2019s own deterministic_accept=false on those pairs, the verdicts should stay more conservative. The 45020 umbrella-canonical framing is also a bit overstated for a mixed cluster." + "accept": true, + "feedback": "The summary is well grounded in the packet, and the accepted soft-edge pairs are conservative and supported by the titles: the docs regression pair, the `_is_hf_initialized` init-empty-weights regression, the Qwen3 ZeRO-3 weight-loading pair, and the training-loss double-shift pair. The false verdicts also read appropriately cautious given only the available evidence." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1864,8 +1825,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -1955,20 +1916,20 @@ "issue:42491|issue:43720", "issue:44661|issue:44805" ], - "prepared_review_unit_hash": "40fa208b96e4260276337f38075fe08501cfab6e109445378f0894e6419f4551", + "prepared_review_unit_hash": "57c57bfc0ad08b2db0944f78527dec842176923827c2dd08b150602b5beecba4", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 35, - "serialized_chars": 38359, + "serialized_chars": 38360, "estimated_input_tokens": 9590, "estimated_eval_tokens": 19436 }, @@ -1977,225 +1938,231 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated reports; only three soft edges are strong enough to accept: 43645\u219443646, 38175\u219443994, and 45672\u219446899. The deepspeed/Zero-3 BertModel pair (43596\u219446515 and 43638\u219446515) is correctly rejected, not treated as a true duplicate cluster.", - "confidence": 0.93, - "canonical_issue_reason": "43646 is the broader, more general report of the Transformers 5 custom-model initialization regression; 43645 is the narrower Jupyter-specific manifestation.", + "summary": "The cluster is mostly heterogeneous: one clear duplicate pair around custom model initialization, one likely duplicate pair around SIGLIP2 output corruption, and many unrelated model-specific regressions.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43632 is the broadest representative of the loading/initialization regression subgroup, centered on the Transformers v5 `_is_hf_initialized` breakage; the other nearby reports are either more specific variants or unrelated.", "canonical_pr_reason": null, - "best_issue_reason": "43646 is the best anchor for the custom-model-init cluster because it describes the core regression without the notebook-specific framing.", + "best_issue_reason": "issue:43632 is the best umbrella issue in this set because it captures the main cross-cutting Transformers v5 model-initialization regression theme, even though the overall cluster contains many unrelated bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Different model families and failure modes: Qwen3.5 PagedAttentionCache/linear_attention vs NemotronH checkpoint loading." + "reason": "Different failures: Qwen3.5 PagedAttentionCache linear_attention crash vs NemotronH checkpoint loading failure." }, { "left": "issue:44863", "right": "issue:46402", "accept": false, - "reason": "Both are loading failures, but they affect different architectures and concrete code paths." + "reason": "Both are loading-related, but one is NemotronH checkpoint loading and the other is Sam3Model text-encoder weights not loading; not the same bug." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF gpt-oss support is unrelated to the Phi-tiny-MoE autocast dtype mismatch." + "reason": "Unrelated: GGUF gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Both are loading-related, but the models and symptoms differ; not the same bug." + "reason": "Both mention tokenizer/model loading, but the affected models and symptoms differ; no clear shared code-path." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "Different Qwen/GGUF support issue vs sliding-window bug; no shared underlying fix." + "reason": "Unrelated: GGUF architecture support vs Qwen3-MoE sliding_window behavior." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Distinct regressions: Qwen2.5-VL rope scaling vs Qwen3_5Moe meta\u2192to_empty materialization." + "reason": "Different Qwen model bugs: temporal position-id scaling vs uninitialized inv_freq after meta materialization." }, { "left": "issue:43596", "right": "issue:46515", "accept": false, - "reason": "The BertModel Zero-3 IndexError is a specific training-time bug, not a general torch 2.4.1 incompatibility." + "reason": "Deepspeed Zero3/BertModel index error is not the same as a general torch 2.4.1 incompatibility report." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Different pipelines and failures: AudioFlamingo3 batched leakage vs Whisper empty-transcription generation." + "reason": "Different audio bugs: track leakage in batched inference vs empty-transcription generation failure." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "GGUF/gpt-oss support is unrelated to the Float8 storage TypeError." + "reason": "Unrelated: GGUF support for gpt-oss vs missing Float8 storage during loading." }, { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Same model family, but different code paths: FSDP2 shared_kv_states rebuild vs device_map=\"auto\" device mismatch." + "reason": "Both are Gemma4-related, but one is FSDP2 shared_kv_states rebuild and the other is device_map forward mismatch; different problems." }, { "left": "issue:43638", "right": "issue:46515", "accept": false, - "reason": "BertModel Zero-3 IndexError is not the same as a broad torch version incompatibility report." + "reason": "Specific Zero3/BertModel index error is not the same as a generic torch version incompatibility." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure vs meta-device loading breakage in different models." + "reason": "Different root issues: Qwen3Moe FP8 failure vs meta-device loading regression." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer corruption is unrelated to Kosmos2.5 OCR indexing." + "reason": "Non-persistent buffer corruption and Kosmos2.5 long-OCR index error are unrelated." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "ImportError for Qwen2.5-VL is unrelated to meta-device loading failures." + "reason": "ImportError for Qwen2_5_VLForConditionalGeneration is unrelated to meta-device loading breakage." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Different symptoms and models: image-preprocessor dtype mismatch vs video StopIteration." + "reason": "Different model families and symptoms: Mistral-3 dtype mismatch vs Qwen3-vl video StopIteration." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Same Transformers 5 custom model initialization regression; 43645 is just the Jupyter-specific variant." + "reason": "Same underlying bug: Transformers 5.0.0 breaks custom model initialization; the notebook-specific report is just a more specific instance of the general issue." }, { "left": "issue:44910", "right": "issue:46858", "accept": false, - "reason": "Both involve attention/cache behavior, but the concrete bugs are different code paths." + "reason": "Both involve Qwen3.5/cache behavior, but one is a Flash Attention position_ids crash and the other is a static-cache slice arithmetic issue; not the same bug." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Qwen3-vl video StopIteration vs MPS histogram support for Int in MoE code." + "reason": "Unrelated: Qwen3-vl embedding video StopIteration vs MPS histogram support for MoE integers." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "3d_parallel runtime issue is unrelated to GGUF gpt-oss support." + "reason": "3d_parallel.py runtime problem is unrelated to GGUF gpt-oss support." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different bugs: torch.compile+SDPA on Bamba vs flash-attention 3D position_ids on Qwen3.5." + "reason": "Different models and code paths: Bamba SDPA torch.compile failure vs Qwen3.5 Flash Attention memory access crash." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Video StopIteration vs TRL SFT unpacking embeddings; different code paths." + "reason": "Both are Qwen-related, but one is video embedding StopIteration and the other is TRL SFT unpacking failure; not the same defect." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import error and Qwen3Moe FP8 failure are unrelated." + "reason": "ImportError for GenerationMixin and Qwen3Moe FP8 failure are unrelated." }, { "left": "issue:44683", "right": "issue:46515", "accept": false, - "reason": "torch>=2.9 compiled flex_attention failure vs torch 2.4.1 incompatibility." + "reason": "Torch>=2.9 flex_attention compilation failure is the opposite end of the version spectrum from a torch 2.4.1 incompatibility report." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Tokenizer regressions, but in different model families and likely different fixes." + "reason": "Both are tokenizer regressions, but on different model families and likely different fixes; not the same underlying bug." }, { "left": "issue:45823", "right": "issue:46515", "accept": false, - "reason": "Gemma4 device mismatch is not the same as a generic torch version incompatibility." + "reason": "Gemma4 PLE device mismatch is unrelated to a generic torch version incompatibility." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "add-new-model-like/tokenizer mapping failure is unrelated to Gemma4 PLE device placement." + "reason": "Add-new-model-like / tokenizer mapping issue is unrelated to Gemma4 device placement during forward." }, { "left": "issue:38175", "right": "issue:43994", "accept": true, - "reason": "Same Siglip2-base-patch16-224 output regression: zero probabilities vs nonsensical results through AutoModel/pipeline." + "reason": "Likely the same SIGLIP2 regression: both report incorrect/nonsensical outputs for google/siglip2-base-patch16-224." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Qwen3-vl video StopIteration vs generic mask/index shape mismatch." + "reason": "Different Qwen3-vl/OCR-related errors with distinct symptoms and no shared concrete code-path." }, { "left": "issue:45672", "right": "issue:46899", - "accept": true, - "reason": "Same Gemma4 4-bit multimodal quantization bug class; both point to unsafe dtype handling in embedders." + "accept": false, + "reason": "Both are Gemma4 multimodal quantization issues, but one is torch.finfo on uint8 audio modules and the other is unsafe dtype casting in embedders; too different to merge." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "_is_hf_initialized flag regression is unrelated to Kosmos2.5 OCR indexing." + "reason": "Broad v5 initialization regression and Kosmos2.5 long-OCR index error are unrelated." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both are config-validation issues, but for different models and without a shared code-path." + "reason": "Both mention config loading/validation, but they affect different model configs and do not clearly share the same bug." }, { "left": "issue:43638", - "right": "issue:44589", + "right": "issue:44661", "accept": false, - "reason": "BertModel Zero-3 IndexError is unrelated to the Float8 storage lookup failure." + "reason": "Zero3/BertModel index error is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different failures: deepspeed zero3 init index error vs new-model helper rejection in tokenizer mappings." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Qwen3_moe LoRA compatibility on hf5.x is unrelated to BitNet packed-weight loading." + "reason": "Qwen3_moe LoRA/hf4.x compatibility and BitNet packed-weight loading are separate issues." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Model-addition/tokenizer-mapping failure is unrelated to the mask shape IndexError." + "reason": "Tokenizer mapping helper failure is unrelated to the shape-mismatch index error." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall conservatism is good, but the summary overstates the evidence by listing the deepspeed/Zero-3 BertModel IndexError as one of the true duplicate clusters. In the packet, both 43596\u219246515 and 43638\u219246515 are explicitly soft-rejected, so that pair should not be described as a true duplicate. The accepted pairs that are grounded are 43645\u219443646, 38175\u219443994, and 45672\u219446899." + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative: it identifies the clear custom-model-init duplicate pair, treats the SIGLIP2 pair as likely rather than certain, and does not overclaim the many unrelated issues as related. The soft-edge rejections are also generally cautious and defensible from the titles alone." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2208,8 +2175,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -2294,229 +2261,229 @@ "issue:44402|issue:44488", "issue:39692|issue:44960" ], - "prepared_review_unit_hash": "5fff046314a9278d71c2d611cb435e47ccdf4f04eb05424eaa223a1026803144", + "prepared_review_unit_hash": "9d6f35dae5ae99e5e94f77991d3ea44700af6df3650b3f7bb90b5dc5d2af412d", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36356, - "estimated_input_tokens": 9089, - "estimated_eval_tokens": 18434 + "serialized_chars": 36359, + "estimated_input_tokens": 9090, + "estimated_eval_tokens": 18436 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Only one soft edge looks like a true duplicate: the two reports of Transformers 5.0.0 filling non-persistent buffers with junk. The rest are related by topic or model family but describe different failures or code paths.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43644 is the clearest representative of the one confirmed duplicate pair, with a broad title that captures the shared regression and earlier report date.", + "summary": "There is one clear duplicate pair in this cluster: the Transformers v5 non-persistent-buffer junk regression. The other soft pairs are conservative rejects because they share only broad themes (model family, loading, quantization, flash-attention, etc.) without the same concrete bug.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43644 is the earlier report of the same non-persistent-buffer junk regression; issue:44534 restates the same symptom with very similar wording.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 best represents the cluster because it names the concrete regression directly and matches the accepted duplicate exactly.", + "best_issue_reason": "issue:43644 is the best representative issue in this set because it cleanly names the confirmed duplicate regression and is the earlier report in the pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Different failures: dequantized model offload/load bug vs Strix Halo mmap OOM." + "reason": "Different problems: dequantized-model load/offload failure vs Strix Halo mmap-related OOM." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Both involve SigLIP2, but one is a zero-probability/model output bug and the other is a broken docs example with processor mismatch and quantization issues." + "reason": "Both mention SigLIP2, but one is an inference/probability bug and the other is a docs example with model/processor mismatch and quantization errors." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Same regression: Transformers v5 filling non-persistent buffers with junk." + "reason": "Same regression: Transformers v5 filling non-persistent buffers with junk; titles and symptom descriptions match closely." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Different model/code path and symptom: Qwen3 auto device-mapping cuda assert vs Qwen3MoE fine-grained FP8 failure." + "reason": "Same broad Qwen3 area, but different concrete failures and code paths (cuda assert on auto device map vs FineGrainedFP8Config failure)." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Unrelated: Qwen3MoE FP8 training/loading issue vs Strix Halo mmap OOM workaround." + "reason": "Unrelated bugs: Qwen3 MoE FP8 config failure vs Strix Halo mmap/OOM workaround." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Different models and mechanisms: Qwen3 MoE hf4.x/hf5.x incompatibility vs Gemma4 resize_token_embeddings propagation bug." + "reason": "Different models and issues: LoRA/qwen3_moe compatibility across branches vs Gemma4 embedding resize propagation." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Unrelated model families and failures." + "reason": "Different model families and failures: GLM5 issue vs Qwen3.5 flash-attention generation crash." }, { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Both are load failures for different checkpoints, but the titles do not show the same concrete bug or shared code path." + "reason": "Both are model-loading failures, but they concern different checkpoints and there is not enough evidence they stem from the same code-path bug." }, { "left": "issue:44863", "right": "issue:46846", "accept": false, - "reason": "Different models and symptoms: NemotronH checkpoint loading vs Qwen3.5 tensor-parallel plan OOM/shape errors." + "reason": "NemotronH checkpoint loading issue vs Qwen3.5 TP-plan missing linear_attn entries; different causes." }, { "left": "issue:46459", "right": "issue:46512", "accept": false, - "reason": "Both are dtype-related, but one is ignored dtype on multimodal loading and the other is stale config dtype after from_config." + "reason": "Both involve dtype handling, but one is AutoModelForCausalLM loading of composite checkpoints and the other is from_config leaving config.dtype stale." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Same family, but different issues: ZeRO-3 load failure for language_model vs incorrect visual encoder keys on save_pretrained." + "reason": "Different Qwen3.5 bugs: ZeRO-3 language_model weight loading vs save_pretrained visual-encoder key regression." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "Unrelated: GraniteMoEHybrid invalid method call vs broken Nemotron example." + "reason": "Different models and failures; no shared code-path bug is evident." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "One is a buffer-initialization regression; the other is a Jupyter/custom-model initialization problem." + "reason": "Different regressions: buffer initialization bug vs Jupyter custom-model initialization breakage." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Same model family, but the first is a broken documentation example and the second reports nonsensical model outputs; not clearly the same bug." + "reason": "Same model family, but one is a broken documentation example and the other is a bad AutoModel/pipeline output regression." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Different models and different loading paths." + "reason": "Different problems: Qwen3 MoE weight conversion with accelerate/deepspeed vs GLM-4.7-Flash loading in tests." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading regressions, but for different models and without evidence of the same underlying tokenizer bug." + "reason": "Both are tokenizer/loading regressions, but they affect different models and likely different tokenizer paths." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Different problems: Qwen3 MoE training memory inefficiency vs quantization/offloading behavior." + "reason": "Different symptoms and mechanisms: Qwen3 MoE training memory inefficiency vs quantization offloading failure." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different models and distinct load failures." + "reason": "Different load-time failures for different models: GLM-4.7-Flash test loading vs Qwen3-VL weight-shape mismatch." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Different model-specific resize_token_embeddings regressions." + "reason": "Unrelated bugs: cross-branch Qwen3 MoE LoRA compatibility vs generic resize_token_embeddings propagation issue." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Both mention memory, but one is training inefficiency and the other is int4 quantization causing reserved-memory OOM." + "reason": "Both mention memory, but one is training inefficiency and the other is int4 quantization reserved-memory OOM." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Different training/optimization bugs with no shared concrete code path." + "reason": "Trainer checkpoint-resume math bug vs optimizer argument handling; different subsystems." }, { "left": "issue:42491", "right": "issue:45335", "accept": false, - "reason": "Different model-specific embedding propagation bugs." + "reason": "Different model-specific bugs: qwen3_moe branch compatibility vs t5gemma embedding resize propagation." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "Unrelated: GLM5 issue vs quantized-model save/load regression." + "reason": "Different issues: GLM5 report vs quantized-model save/load breakage with sinq." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Both are save/load/init issues, but the concrete failures differ: extra_state handling vs unexpected _is_hf_initialized argument." + "reason": "Both affect save/load, but one is extra_state handling and the other is init_empty_weights/_is_hf_initialized compatibility." }, { "left": "issue:46897", "right": "issue:46901", "accept": false, - "reason": "Similar loss-shift pattern, but different model implementations; not clearly the same underlying bug." + "reason": "Similar training-loss-shift pattern, but they are different model implementations and not clearly the same concrete bug." }, { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and describe different incorrect behaviors." + "reason": "Different tokenizer regressions in different model families; not the same underlying bug." }, { "left": "issue:45276", "right": "issue:45335", "accept": false, - "reason": "Similar API symptom, but different affected submodules and model implementations; not enough to treat as the same bug." + "reason": "Similar API surface, but different model architectures and embedding components; insufficient evidence of one shared bug." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Unrelated failures in different components/models." + "reason": "Different failures: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer in serve." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Different flash-attention crashes with different root causes and error modes." + "reason": "Both touch flash_attention, but one is a packed-sequence/position_ids crash and the other is an s_aux=None AttributeError." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Related to tokenizer/model loading, but one is vocab-size mismatch and the other is a model-load regression for a different checkpoint." + "reason": "Different models and symptoms: vocab-size/token mismatch vs model loading failure." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Unrelated model families and failure modes." + "reason": "Unrelated model families and bugs; only superficial similarity in being model-specific regressions." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the single accepted soft edge (issue:43644 \u2194 issue:44534) is well supported by the near-identical titles. The other pairwise rejections are conservative and generally stick to differences visible in the issue titles. No obvious overstated duplicate claim." + "accept": false, + "feedback": "The core duplicate call on issue:43644 \u2194 issue:44534 is grounded, but the canonical_issue_reason overstates the evidence by calling 43644 the 'more complete' report. In the packet, 44534 has the longer body length, so completeness is not supported here. The rest of the soft-edge rejections read conservative enough, but this unsupported comparative claim means I can\u2019t accept the summary as written." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2529,8 +2496,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -2615,226 +2582,226 @@ "issue:43646|issue:43708", "issue:44589|issue:44756" ], - "prepared_review_unit_hash": "e4233b2de85ab8ae5f65f7ae00ac780ada1a2fa1227468245cbce5283b0fd3ab", + "prepared_review_unit_hash": "24780ce2d5060a6333a1dd156e862c60b0ded63649f46703cdccb8545ff9aae7", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36537, - "estimated_input_tokens": 9135, - "estimated_eval_tokens": 18526 + "serialized_chars": 36536, + "estimated_input_tokens": 9134, + "estimated_eval_tokens": 18524 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated transformer issues, with one small duplicate cluster around `resize_token_embeddings` not propagating to model output/decoder embeddings across Gemma4/T5Gemma variants.", - "confidence": 0.9, - "canonical_issue_reason": "Issue 45292 is the broadest, most generic report of the embedding-resize propagation bug and best serves as the umbrella duplicate target for the model-specific variants.", + "summary": "Most candidate pairs are only loosely similar and should remain separate. The only clear duplicate chain is the `resize_token_embeddings` regression affecting output/input embeddings across model variants; everything else looks like distinct bugs.", + "confidence": 0.78, + "canonical_issue_reason": "Issue 45292 is the broadest, most generic report of the shared `resize_token_embeddings` bug, and it cleanly subsumes the model-specific variants in 45276 and 45335.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45292 is the clearest representative of the shared bug class: `resize_token_embeddings` updates are not reflected in downstream embeddings/output weights, with nearby issues describing the same failure in specific model families.", + "best_issue_reason": "Issue 45292 is the best representative of the only convincing duplicate cluster here because it states the underlying bug without model-specific narrowing.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "One is a Torch version compatibility complaint; the other is a meta-device loading regression affecting specific models. Different bugs and fixes." + "reason": "Both are loading-related, but one is a Torch compatibility failure and the other is a `torch.device(\"meta\")` model-loading regression; different root causes." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert vs AudioFlamingo3 batch token/embedding leakage. Not the same code path." + "reason": "CUDA assert during auto device mapping is unrelated to AudioFlamingo batched inference token/embedding leakage." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Gemma4 FSDP2 shared-kv state error vs Qwen multimodal config unwrapping in `from_config`. Separate model-initialization problems." + "reason": "Gemma4 FSDP2 `KeyError` and Qwen multimodal `text_config` unwrapping are different code paths and different failures." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both mention untied `lm_head`, but they are different model implementations and likely different missing tie-weight fixes." + "reason": "Both mention lm_head tying, but they are model-specific reports with different affected architectures and likely different fixes." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "`init_empty_weights` argument breakage is unrelated to the Strix Halo mmap/OOM issue." + "reason": "`init_empty_weights` argument handling and Strix Halo mmap/OOM are unrelated problems." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing with PEFT LoRA is unrelated to the T5 apex RMSNorm memory leak." + "reason": "Gradient checkpointing with PEFT LoRA is a different bug from T5/apex RMSNorm memory leak." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both involve memory/quantization, but one is offloading behavior and the other is increased reserved memory under int4 quantization. Not the same defect." + "reason": "Both involve memory/quantization, but one is offloading behavior and the other is increased reserved CUDA memory under int4." }, { "left": "issue:44079", "right": "issue:46829", "accept": false, - "reason": "ModelOutput key assignment bug vs VideoPrism output field mismatch. Different output-structure issues." + "reason": "ModelOutput key assignment logic is unrelated to VideoPrism's incorrect `hidden_states` field mapping." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Gemma4 FSDP2 attention state bug vs Qwen3.5 MoE rotary embedding CPU-offload incompatibility. Unrelated." + "reason": "Gemma4 FSDP2 attention state bug and Qwen3_5Moe CPU-offload incompatibility are separate issues." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Distinct quantization/model-loading failures for different models and backends." + "reason": "`init_empty_weights` loading error and MXFP4 quantization fallback are different model-loading regressions." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "Import path breakage for `GenerationMixin` is unrelated to Whisper pipeline `return_language` regression." + "reason": "Importing `GenerationMixin` and Whisper `return_language` pipeline behavior are unrelated." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Same broad area of gradient checkpointing, but one is ineffective LoRA checkpointing and the other is eval-mode prohibition. Different problems." + "reason": "One is ineffective gradient checkpointing with LoRA; the other is checkpointing not allowed in eval mode." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "AudioFlamingo3 batched inference leak is unrelated to Gemma4 device-map mismatch during forward." + "reason": "AudioFlamingo batching leak and Gemma4 PLE device mismatch are not the same underlying bug." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Int4 quantization OOM from reserved memory is not the same as disabling mmap on Strix Halo." + "reason": "Int4 quantization OOM and Strix Halo mmap OOM are both memory-related but arise from different mechanisms." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Meta-device model loading failure and Strix Halo mmap/OOM are different memory/loading regressions." + "reason": "Meta-device loading breakage and Strix Halo mmap disablement target different failure modes." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "SwitchTransformers sparse-layer construction bug vs Gemma4 FSDP2 shared-kv-state error. Different model subsystems." + "reason": "SwitchTransformers sparse-layer config and Gemma4 FSDP2 attention state rebuilding are unrelated." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "NemotronH checkpoint loading failure vs broken example script. Related family, but not the same concrete bug." + "reason": "NemotronH checkpoint loading and the broken official Nemotron example are separate reports, not one concrete bug." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same underlying embedding-resize propagation bug; 45276 is a Gemma4-specific manifestation of 45292's generic issue." + "reason": "Both describe the same `resize_token_embeddings` propagation failure; 45276 is the Gemma4-specific manifestation of the broader 45292 bug." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Video input regression for Qwen VL models vs strict config rejection for granite_speech. Different failures." + "reason": "Qwen VL video-input regression and strict config loading for granite_speech are different subsystems and different bugs." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Eval-mode gradient checkpointing prohibition vs RLHF NaN from 3D position_ids/SDPA fallback. Not the same bug." + "reason": "Eval-mode checkpointing restriction is unrelated to RLHF NaNs from 3D position_ids and SDPA fallback." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "save/from_pretrained `extra_state` handling is unrelated to the Strix Halo mmap issue." + "reason": "`extra_state` save/load handling and Strix Halo mmap OOM do not share the same root cause." }, { "left": "issue:42796", "right": "issue:46326", "accept": false, - "reason": "Tokenizer BPE-marker handling and `continue_final_message` behavior are separate tokenizer bugs." + "reason": "Tokenizer BPE-marker handling and `continue_final_message` behavior are both tokenizer-adjacent but distinct bugs." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Flash-attention crash on `s_aux=None` is unrelated to Gemma4 FSDP2 shared state rebuilding." + "reason": "Flash-attention `s_aux=None` crash and Gemma4 FSDP2 KeyError are separate implementation issues." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Model loading test failure for GLM-4.7-Flash vs Qwen3.5 save_pretrained regression. Different code paths." + "reason": "Unit-test loading of GLM-4.7-Flash and Qwen3.5 save_pretrained regression are unrelated." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Both concern custom-model save/load, but one is permissions propagation and the other is wrong module reloading. Not duplicates." + "reason": "Read-only permission propagation from `save_pretrained` is a different issue than loading the wrong custom module after saving." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same resize-token-embedding propagation bug; 45335 is a T5Gemma-specific instance of the broader issue in 45292." + "reason": "Both are the same embedding-resize regression: 45292 is the generic output-embedding report and 45335 is the decoder-embedding variant." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Jupyter/custom-model initialization breakage is separate from `extra_state` save/load handling." + "reason": "Custom model initialization in notebooks is unrelated to `extra_state` save/from_pretrained handling." }, { "left": "issue:42175", "right": "issue:43957", "accept": false, - "reason": "Packaging/backend dependency omission is unrelated to meta-device loading regressions." + "reason": "Missing TensorFlow in the torch extra and meta-device model-loading failures are unrelated." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "Different model-tokenizer loading issues; one is vocab/token placement, the other is loading a specific model in latest version." + "reason": "Tokenizer vocab-size mismatch and ScandiBERT loading failure are different bugs." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization regression vs Trainer resume/max_steps miscalculation. Different functionality." + "reason": "Custom model initialization regression and `resume_from_checkpoint` max_steps calculation are unrelated." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage deserialization error is unrelated to Strix Halo mmap/OOM." + "reason": "Float8 storage lookup failure and Strix Halo mmap/OOM are different low-level problems." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only clear cluster is the resize_token_embeddings propagation issues around issue:45292, issue:45276, and issue:45335, and the rejection calls for the other soft pairs are appropriately cautious based on the titles alone. No evidence of overstated duplicate claims or mergeability issues." + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The only supported duplicate cluster is the resize_token_embeddings trio around 45292/45276/45335, and the other pair rejections are appropriately cautious." }, "evaluator_used": true, "retried": false, @@ -2850,8 +2817,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -2938,226 +2905,226 @@ "issue:43632|issue:47050", "issue:43596|issue:44683" ], - "prepared_review_unit_hash": "1efd394f368059374da1edb2ba6255643322031621577b41491ec2e07e0d9ba7", + "prepared_review_unit_hash": "2d55080040718aa84e31238e8de0321881a572fac0eb3ee92f0dbf0e3b2571d7", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37524, - "estimated_input_tokens": 9381, - "estimated_eval_tokens": 19018 + "serialized_chars": 37511, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers regressions and docs issues; the soft pairs are similar at a surface level but do not share the same concrete bug or fix path.", - "confidence": 0.89, - "canonical_issue_reason": "issue:45701 is the broadest representative of the recurring tokenizer-regression theme in this set, with a title that most naturally covers version-dependent tokenization changes.", + "summary": "This cluster is dominated by unrelated Transformers v5 regression reports across different model families and subsystems. The soft pairs mostly share only broad symptom words (tokenization, dtype mismatch, loading, docs, or training loss), not a clearly identical bug or patch target, so they should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "Issue 45812 is the strongest representative because it is a concrete, high-signal v5 regression with a clear user-facing failure mode and broad impact within one subsystem (tokenization), unlike the more generic or model-specific reports elsewhere in the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45701 is the most suitable global representative because it is broad, regression-oriented, and closest to the common symptom space across several other tokenizer-related reports.", + "best_issue_reason": "Issue 45812 is the best overall issue to anchor this cluster: it is specific enough to be actionable, clearly describes a regression, and is more representative than vague umbrella reports or one-off model loading errors.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46894", "right": "issue:46901", "accept": false, - "reason": "Both mention a training-loss shift pattern, but they are model-specific reports with no evidence of the same shared code path." + "reason": "Both describe a training-loss label-shift bug, but they are for different model families and lack evidence of the same shared implementation path." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "One is a general Torch 2.10 incompatibility report; the other is a torch.autocast dtype mismatch in Phi-tiny-MoE. Different failures." + "reason": "One is a Torch compatibility complaint, the other is a dtype mismatch in a specific model under autocast; the failure modes and code paths are different." }, { "left": "issue:46315", "right": "issue:46491", "accept": false, - "reason": "Performance regression in convert_tokens_to_ids vs. a leading-space round-trip bug. Different tokenizer behaviors." + "reason": "Performance regression in slow tokenizer conversion vs a CodeLlama encode/decode spacing bug; these are unrelated tokenizer issues." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Wrong token IDs for Granite models is unrelated to added-token lookup slowdown." + "reason": "Wrong token IDs for Granite models is a correctness regression; slow convert_tokens_to_ids is a performance regression, not the same bug." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language failure is a pipeline behavior bug, not the same issue as general tokenization changes." + "reason": "Whisper pipeline return_language and a generic tokenization-change report are not the same concrete failure." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "PagedAttentionCache linear_attention crash and a template-node compilation TypeError are different code paths." + "reason": "PagedAttentionCache linear_attention crash and a compiler template-node error are different subsystems and different root causes." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Both are v5 breakages, but one is base_model_prefix loading and the other is notebook-defined custom model initialization." + "reason": "Both involve v5 loading behavior, but one is base_model_prefix loading and the other is custom model init in notebooks; no shared concrete bug is shown." }, { "left": "issue:43425", "right": "issue:46869", "accept": false, - "reason": "Torch version incompatibility is unrelated to YOLO multi-GPU device_map loading." + "reason": "Torch incompatibility and multi-GPU device-map failure are different issues; they only overlap at a high level." }, { "left": "issue:39404", "right": "issue:46665", "accept": false, - "reason": "Whisper pipeline return_language and from_pretrained-inside-classes are distinct behaviors with different root causes." + "reason": "Whisper pipeline return_language and from_pretrained-inside-classes describe different call paths and different symptoms." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "flex_attention compilation on torch>=2.9 is unrelated to Gemma4 device placement during forward." + "reason": "Compiled flex_attention failure on torch>=2.9 is unrelated to Gemma4 device mismatch in forward." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "General Torch compatibility and Mistral-3 image-preprocessor dtype mismatch are not the same bug." + "reason": "Torch version incompatibility is not the same as the Mistral-3 image preprocessor/model dtype mismatch." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "These are both compatibility-style reports, but they affect different models and different failure modes." + "reason": "These are both broad compatibility complaints, but they target different models and distinct failures." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video-input regression for Qwen VL models does not match a generic mask-shape IndexError." + "reason": "Both are regression reports, but one is a video-input model family regression and the other is a tensor-mask shape error; not the same bug." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to the non-template-node compilation error." + "reason": "Float8 storage lookup failure and compiler template-node errors are unrelated runtime/compiler problems." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs-related, but they target different pipeline task deprecations and behavior changes." + "reason": "Both are docs-related, but they discuss different pipeline/task deprecations and do not point to one underlying issue." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "An isolated reproduction of another issue is not the same as the PagedAttentionCache crash." + "reason": "An isolated reproduction of an older issue and a PagedAttentionCache crash on Qwen3.5 are different problems." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Wrong/incoherent model outputs for SigLIP2 are not the same as ModelOutput key assignment semantics." + "reason": "Wrong SigLIP2 outputs and incorrect ModelOutput key assignment are unrelated correctness bugs." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export failures for torch_compilable_check are unrelated to Perceiver non-default-resolution failures." + "reason": "torch.export failures via torch_compilable_check and Perceiver interpolation failures involve different code paths and different triggers." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "T5Tokenizer add_special_tokens conflict and gguf_file loading breakage are separate tokenizer-loading bugs." + "reason": "Tokenizer add_special_tokens conflict and gguf_file from_pretrained breakage are separate loading issues." }, { "left": "issue:46050", "right": "issue:46326", "accept": false, - "reason": "Sinq save/load breakage and mistral_common continue_final_message behavior are different features." + "reason": "Quantization save/load breakage and continue_final_message behavior in mistral_common tokenizers are unrelated." }, { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "extra_state save/load handling is unrelated to unwrapping text_config in composite multimodal configs." + "reason": "extra_state save/load handling and from_config text_config unwrapping are different serialization/configuration bugs." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper return_language and MXFP4 quantization fallback are different regressions." + "reason": "Whisper language return and MXFP4 quantization fallback are unrelated regressions." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "Reproduction of a prior issue and a tokenizer message-continuation bug do not share the same underlying defect." + "reason": "These refer to different tokenizer behaviors and different model stacks; no shared defect is demonstrated." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT masked-LM prediction regression is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + "reason": "CamemBERT MLM prediction errors and tokenizer loading failures for EMBEDDIA/est-roberta are different issues." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Both involve tokenizer loading, but one is a Mistral regex patch crash and the other is a different model's tokenizer load failure." + "reason": "Both are tokenizer-loading regressions, but one is a tokenizer file load issue and the other is a Mistral regex patch crash; not the same bug." }, { "left": "issue:45823", "right": "issue:46899", "accept": false, - "reason": "Gemma4 device-map placement issues and Gemma4 multimodal 4-bit dtype-cast issues are related only by model family, not the same bug." + "reason": "Gemma4 device mismatch in forward and Gemma4 multimodal 4-bit dtype-cast breakage are distinct failures." }, { "left": "issue:45479", "right": "issue:46894", "accept": false, - "reason": "A sequence-classification loss degeneracy is not the same as ASR training-loss double-shift behavior." + "reason": "Zero-loss classification bug and training-loss double-shift bug are both loss-related but not the same underlying defect." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Qwen3 MoE training memory usage and Strix Halo mmap OOM are different resource problems." + "reason": "Qwen3 MoE training memory inefficiency and Strix Halo mmap OOM are separate performance/memory issues." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Qwen3VL MoE loading regression and Qwen3-VL weight-shape mismatch are separate loading failures." + "reason": "Qwen3VL MoE loading breakage and Qwen3-VL weight-shape mismatch are both model-loading issues, but with different concrete errors." }, { "left": "issue:44164", "right": "issue:46821", "accept": false, - "reason": "extra_state serialization and a TimesFM window_size attribute error are unrelated." + "reason": "extra_state save/load handling and TimesFM window_size AttributeError are unrelated issues." }, { "left": "issue:43632", "right": "issue:47050", "accept": false, - "reason": "_is_hf_initialized flag breakage and Canine ONNX fp16 dtype mismatch are different code paths." + "reason": "_is_hf_initialized flag regression and CanineModel fp16 ONNX export dtype mismatch are different bug classes." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "DeepSpeed ZeRO-3/BertModel index error is unrelated to compiled flex_attention failures." + "reason": "Deepspeed ZeRO-3/BertModel index error and compiled flex_attention torch>=2.9 failure are unrelated." } ] }, @@ -3176,8 +3143,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -3262,20 +3229,20 @@ "issue:43317|issue:43482", "issue:43746|issue:46326" ], - "prepared_review_unit_hash": "d4d5ee1970d314d415e63770e7f0a8bc930971981f762aa0a7d4d6be33932315", + "prepared_review_unit_hash": "e128507d98f0b61cd5c14a15216d10a2ef8d5aa04560b0f3350029a19252e4e1", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36749, + "serialized_chars": 36751, "estimated_input_tokens": 9188, "estimated_eval_tokens": 18632 }, @@ -3284,202 +3251,190 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue set with no PRs. The one audio-feature-extraction pair is only superficially related from titles/metadata, so it should not be merged as the same bug on the available evidence.", - "confidence": 0.66, - "canonical_issue_reason": "Issue 44206 is the clearest standalone issue in the audio-feature-extraction area, but the metadata is not strong enough to prove 44220 is the same underlying defect.", + "summary": "The set is mostly a grab bag of unrelated regression reports. The only plausible duplicate pair is the GGUF-loading regression on v5, where one issue is the generic breakage and the other is a model-specific manifestation. No PRs are present.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43452 is the broadest report of the GGUF loading regression, affecting both AutoTokenizer and AutoModelForCausalLM; it is the best umbrella issue for the related Qwen2.5-GGUF failure.", "canonical_pr_reason": null, - "best_issue_reason": "44206 is the best issue anchor because it names a concrete regression and failing path more clearly than 44220.", + "best_issue_reason": "issue:43452 is the strongest canonical issue candidate because it describes the underlying loading-path breakage in a model-agnostic way and can subsume the more specific GGUF symptom report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45479", "right": "issue:46897", "accept": false, - "reason": "Different bug classes: zero-loss classification vs Florence2 training double-shift." + "reason": "Different problems: one is a generic zero-loss classification regression, the other is a Florence2 training-loss double-shift bug." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are load failures, but they involve different models and no shared concrete code path." + "reason": "Both are loading regressions, but they affect different models and likely different tokenizer/load paths." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Repository tokenizer loading and Mistral regex-patching are separate tokenizer paths." + "reason": "Different tokenizer-loading failures: ScandiBERT load failure vs a Mistral regex patch crash." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Base-model-prefix loading breakage and non-persistent buffer junk are distinct regressions." + "reason": "Both are v5 regressions, but one is about base_model_prefix loading and the other about junk in non-persistent buffers." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Tokenizer BPE-marker handling is unrelated to tied lm_head weights in Mistral3." + "reason": "Unrelated issues: tokenizer BPE marker handling vs lm_head weight tying in Mistral3." }, { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Repository tokenizer.json handling and DeepSeek-Coder tokenizer output are different bugs." + "reason": "Both involve tokenizers, but one is about AutoTokenizer ignoring tokenizer.json and the other is a DeepSeek-Coder output regression." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "T5 tokenizer method conflict and Qwen3Moe FP8 config failure are unrelated." + "reason": "Completely different areas: T5 tokenizer method conflict vs Qwen3Moe FP8 config failure." }, { "left": "issue:46285", "right": "issue:46402", "accept": false, - "reason": "CLIP structure mismatch and Sam3 text-encoder weight loading affect different model code paths." + "reason": "Different model-architecture regressions: CLIP structure mismatch vs Sam3 text encoder weights not loading." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Image preprocessor dtype mismatch is unrelated to template-node compilation errors." + "reason": "Different codepaths: Mistral multimodal dtype mismatch vs a compilation/template-node error." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper return_language pipeline issue and CamemBERT masked-LM predictions are not the same bug." + "reason": "Different bugs in different models: Whisper pipeline return_language vs CamemBERT masked LM predictions." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Tokenizer mapping registration failure and flex_attention compile regression are unrelated." + "reason": "No shared underlying bug: model registration failure vs torch flex_attention compilation on newer torch." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "trust_remote_code field loss and PagedAttentionCache group-type crash are different issues." + "reason": "Trust-remote-code config field loss is unrelated to PagedAttentionCache's linear_attention crash." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "XLNet CPU arange placement and Qwen offload incompatibility are separate device bugs." + "reason": "Both are device/offload-related, but they target different models and different failure mechanisms." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Qwen3Moe FP8 config failure and PagedAttentionCache invalid-group-type crash do not share a concrete fix." + "reason": "Same broad Qwen family, but distinct concrete failures: FP8 config vs PagedAttentionCache group-type crash." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Both are load failures after updates, but they target different models and symptoms." + "reason": "Both are model-loading regressions, but they are different models and not clearly the same underlying defect." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "High perplexity on one model is unrelated to tokenizer post-processor update behavior." + "reason": "Perplexity regression on a model is unrelated to runtime post-processor updates for special tokens." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Torch dynamo grouped_mm failure and non-template-node compile error are different failures." + "reason": "Different failures: torch dynamo grouped_mm tracing error vs template-node compilation error." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "Trainer token counting bug and multi-image IterableDataset finetuning failure are unrelated." + "reason": "Trainer token-count tracking and multimodal iterable-dataset failures are unrelated." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Non-persistent buffer corruption is not the same concrete bug as generic custom model initialization breakage." + "reason": "Both are v5 custom-model regressions, but one concerns buffers and the other model initialization behavior." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Qwen2.5-VL multi-image finetuning and mistral_common continue_final_message are unrelated." + "reason": "Different subsystems: multimodal fine-tuning with iterable datasets vs mistral_common tokenizer behavior." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError and _is_hf_initialized regression are different issues." + "reason": "AppleSilicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "TensorFlow extra packaging and backend mapping validation are separate backend problems." + "reason": "Packaging/backend availability issues, but different backends and different failure points." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "The titles both concern audio feature extraction, but 44206 is a specific unsupported-`center` regression while 44220 only names `_torch_extract_fbank_features()`; titles/metadata do not establish the same concrete code-path bug." + "reason": "Related audio area, but one is a feature-extractor arg bug and the other is a lower-level fbank feature issue; not clearly the same defect." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Tokenizer special-token method conflict and notebook custom-model initialization are unrelated v5 regressions." + "reason": "Tokenizer special-token conflict is unrelated to Jupyter custom model initialization." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Whisper skip_special_tokens decoding and empty-transcription generation after align_special_tokens are different behaviors." + "reason": "Both are Whisper special-token bugs, but one is decode-API behavior and the other is generation failure on empty transcription." }, { "left": "issue:43452", "right": "issue:43482", - "accept": false, - "reason": "Generic gguf_file from_pretrained breakage and Qwen2.5-GGUF loading failure are not the same bug." - }, - { - "left": "issue:44987", - "right": "issue:44991", - "accept": false, - "reason": "Both are versioned tokenizer-load failures, but the models and failure modes differ." - }, - { - "left": "issue:45084", - "right": "issue:45507", - "accept": false, - "reason": "Compilation/template-node error and GraniteMoEHybrid invalid-method call are unrelated." + "accept": true, + "reason": "Same underlying GGUF loading regression on v5: the generic `gguf_file` breakage matches the Qwen2.5-GGUF load failure." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "GPU+CPU offload dequantized-load failure and GGUF model loading failure are different code paths." + "reason": "Different offload/loading problems: dequantized model device_map handling vs Qwen2.5-GGUF version-specific load failure." }, { "left": "issue:43746", "right": "issue:46326", "accept": false, - "reason": "PEFT checkpoint loading and mistral_common continue_final_message are unrelated." + "reason": "Different model families and different load/runtime paths; no same-bug evidence." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overview is mostly conservative, but it overstates the evidence for the one accepted pair. In the packet we only have titles and metadata for 44206/44220, which show related audio-feature-extraction symptoms, not a clearly established shared crash path or identical function-level bug. That makes the acceptance too assertive for the available evidence. The rest of the rejections are generally conservative, and there are no PRs to assess." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative about the soft edges. The only accepted duplicate pair, issue:43452 vs issue:43482, is reasonably supported by the titles as a generic GGUF loading regression versus a model-specific manifestation. The rest of the verdicts appropriately avoid over-claiming similarity, and there are no PR pairs to assess." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3491,8 +3446,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -3575,214 +3530,208 @@ "issue:43531|issue:44877", "issue:36032|issue:41720" ], - "prepared_review_unit_hash": "3f4061866955ef529d56ea59792d5b37e315679b2cebd471098b614684e5f766", + "prepared_review_unit_hash": "53e3b227ebbee2bae89fb56aeb9a9bcd75629b08d7feb5ed72abb82bc1771b3e", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35993, - "estimated_input_tokens": 8999, - "estimated_eval_tokens": 18254 + "serialized_chars": 36004, + "estimated_input_tokens": 9001, + "estimated_eval_tokens": 18258 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a loose mix of unrelated Transformers issues; there is no PR data. The closest thematic concentration is around v5 initialization/regression problems, but none of the soft edges look like true duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43632 is the strongest anchor because it states a broad v5 initialization regression with a likely root-cause flag (`_is_hf_initialized`) that overlaps conceptually with several nearby initialization-related reports.", + "summary": "The cluster is a collection of mostly unrelated Transformers bug reports; the soft pairs are mostly same-surface-area but not the same underlying defect, so they should stay separate.", + "confidence": 0.65, + "canonical_issue_reason": "Issue 43646 is the broadest, most central regression-style report in the set: Transformers v5 breaking custom model initialization. It is a reasonable anchor for the cluster\u2019s general v5/model-init breakage theme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43632 is the best representative issue for this cluster: broad, root-cause-oriented, and closer to the shared v5 regression theme than the more model-specific reports.", + "best_issue_reason": "Issue 43646 is the best standalone representative because it describes a general, high-impact API break rather than a model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43262 \u2014 audio processor chat template defaults to 16kHz", - "right": "issue:43377 \u2014 MIMI encoder batched vs single outputs differ", + "left": "issue:43262", + "right": "issue:43377", "accept": false, - "reason": "Both are audio-related, but one is about sampling-rate defaults and the other about padding-mask/batching behavior in a different model." + "reason": "Both are audio-related, but one is a chat-template sampling-rate default bug and the other is a batching/padding-mask encoder mismatch." }, { - "left": "issue:43208 \u2014 xLSTM bugs preventing training <7B", - "right": "issue:47013 \u2014 xLSTM train crash when return_last_states=False", + "left": "issue:43208", + "right": "issue:47013", "accept": false, - "reason": "Same family, but the titles describe different defects and code paths; not the same underlying bug." + "reason": "Both involve xLSTM training, but they describe different failures: multiple bugs for small models vs a train-mode unpacking crash." }, { - "left": "issue:43883 \u2014 Molmo missing all_tied_weights_keys", - "right": "issue:44488 \u2014 cannot load cjvt/sleng-bert", + "left": "issue:43883", + "right": "issue:44488", "accept": false, - "reason": "These are unrelated loading failures with different models and different failure modes." + "reason": "Both are model-loading errors, but they affect different models and different code paths." }, { - "left": "issue:43792 \u2014 whisper-large-v2 can't run", - "right": "issue:46869 \u2014 YOLO model failing on multi-GPU auto device map", + "left": "issue:43792", + "right": "issue:46869", "accept": false, - "reason": "Different models, different execution paths, and different error types." + "reason": "No concrete overlap beyond generic runtime failure; different models and different device-mapping behavior." }, { - "left": "issue:44403 \u2014 unnecessary noise when loading a transformer", - "right": "issue:45081 \u2014 Mistral tokenizer regex patch crashes", + "left": "issue:44403", + "right": "issue:45081", "accept": false, - "reason": "One is cosmetic logging noise; the other is a tokenizer crash. Not the same bug." + "reason": "One is about noisy loading messages, the other is a tokenizer regex patch crash; not the same bug." }, { - "left": "issue:43646 \u2014 Transformers 5.0.0 breaks custom model initialization", - "right": "issue:44534 \u2014 v5 fills non-persistent buffers with junk", + "left": "issue:43646", + "right": "issue:44534", "accept": false, - "reason": "Both are v5 regressions, but they affect different mechanisms: custom init vs buffer state handling." + "reason": "Both are v5 regressions, but one is custom model initialization and the other is non-persistent buffer serialization corruption." }, { - "left": "issue:43645 \u2014 Jupyter notebook custom model init breaks in v5", - "right": "issue:46661 \u2014 AutoModel.from_pretrained remote-code config_class missing", + "left": "issue:43645", + "right": "issue:46661", "accept": false, - "reason": "Both involve model loading/init, but the root causes and affected flows differ." + "reason": "Both concern custom/remote-code model initialization, but the failures are different: notebook-defined custom models vs missing config_class on remote-code models." }, { - "left": "issue:42915 \u2014 Qwen3Moe fails with FineGrainedFP8Config", - "right": "issue:45084 \u2014 Can't compile non template nodes", + "left": "issue:42915", + "right": "issue:45084", "accept": false, - "reason": "Different subsystems: quantization/config handling versus compilation/template-node parsing." + "reason": "Different subsystems: fine-grained FP8 Qwen3Moe failure vs a template compilation TypeError." }, { - "left": "issue:44849 \u2014 Qwen3.5 output_hidden_states bug", - "right": "issue:45325 \u2014 Qwen2.5-VL still-image temporal position_ids scaled wrongly", + "left": "issue:44849", + "right": "issue:45325", "accept": false, - "reason": "Different models and different bugs in output handling vs rope/position-id computation." + "reason": "Both are Qwen3.x multimodal regressions, but one affects hidden states and the other rope-index scaling for still images." }, { - "left": "issue:43377 \u2014 MIMI batching/padding-mask mismatch", - "right": "issue:43819 \u2014 DAC.from_latents mismatch with forward pass", + "left": "issue:43377", + "right": "issue:43819", "accept": false, - "reason": "Similar symptom class, but different models and different implementation details; not the same underlying defect." + "reason": "Both involve audio codecs, but one is padding-mask batching and the other is DAC latent decoding mismatch." }, { - "left": "issue:43404 \u2014 lm_head weight not tied in Mistral3ForConditionalGeneration", - "right": "issue:44466 \u2014 lm_head.weight serialization depends on device in v5", + "left": "issue:43404", + "right": "issue:44466", "accept": false, - "reason": "Both touch tied weights, but one is a model-architecture tying bug and the other is serialization inconsistency." + "reason": "Both mention tied lm_head weights, but one is an actual model tying bug and the other is device-dependent serialization inconsistency." }, { - "left": "issue:44987 \u2014 transformers>=5.1.0 fails loading physical-intelligence/fast", - "right": "issue:45081 \u2014 Mistral tokenizer regex patch crashes", + "left": "issue:44987", + "right": "issue:45081", "accept": false, - "reason": "Both are loading problems, but they target different components and causes." + "reason": "Different loading failures: one is a version-specific model load failure, the other is a Mistral tokenizer regex patch crash." }, { - "left": "issue:28530 \u2014 early stopping needs metric_for_best_model but lacks eval_f1", - "right": "issue:34567 \u2014 num_input_tokens_seen not updating", + "left": "issue:28530", + "right": "issue:34567", "accept": false, - "reason": "Completely different trainer-state behaviors with no shared failure mode." + "reason": "Completely unrelated training-state bugs: early stopping metric selection vs token-count tracking." }, { - "left": "issue:43994 \u2014 siglip2 gives nonsensical AutoModel/pipeline results", - "right": "issue:44610 \u2014 OmDet-Turbo processor outputs 640px instead of 224px", + "left": "issue:43994", + "right": "issue:44610", "accept": false, - "reason": "Both involve bad outputs, but one is model/pipeline behavior and the other is a processor image-size mismatch." + "reason": "Different components and symptoms: bad AutoModel/pipeline outputs vs processor image-size mismatch." }, { - "left": "issue:44464 \u2014 chunked generation inconsistent with compiled forward", - "right": "issue:45987 \u2014 StaticCache.get_seq_length returns Tensor instead of int", + "left": "issue:44464", + "right": "issue:45987", "accept": false, - "reason": "Both are generation/cache-related, but the concrete bugs are different." + "reason": "One is chunked-generation inconsistency with compiled forward; the other is a cache-length return-type contract issue." }, { - "left": "issue:42915 \u2014 Qwen3Moe FineGrainedFP8Config failure", - "right": "issue:44291 \u2014 init_empty_weights passes unexpected _is_hf_initialized", + "left": "issue:42915", + "right": "issue:44291", "accept": false, - "reason": "Different regression areas: FP8 quantization versus initialization API incompatibility." + "reason": "Both touch loading/initialization, but one is Qwen3Moe FP8 and the other is init_empty_weights/_is_hf_initialized compatibility." }, { - "left": "issue:44451 \u2014 latest version cannot load ScandiBERT", - "right": "issue:45081 \u2014 Mistral tokenizer regex patch crashes", + "left": "issue:44451", + "right": "issue:45081", "accept": false, - "reason": "Both are model loading issues, but the models and causes are unrelated." + "reason": "Different tokenizer-loading failures; no evidence they share the same root cause." }, { - "left": "issue:36032 \u2014 T5Tokenizer add_special_tokens method conflict", - "right": "issue:43742 \u2014 KeyError loading facebook/MobileLLM-125M", + "left": "issue:36032", + "right": "issue:43742", "accept": false, - "reason": "Tokenizer/loading failures, but not the same bug or code path." + "reason": "Different tokenizer/model issues: T5 tokenizer API conflict vs a key error loading a specific model." }, { - "left": "issue:44008 \u2014 Gemma3n variable name collision in forward", - "right": "issue:46899 \u2014 Gemma4 multimodal 4-bit quantization dtype bug", + "left": "issue:44008", + "right": "issue:46899", "accept": false, - "reason": "Same broader product area, but distinct bugs: name collision vs unsafe dtype casting." + "reason": "Both are Gemma multimodal issues, but one is a forward variable collision and the other is unsafe dtype casting under 4-bit quantization." }, { - "left": "issue:44265 \u2014 torch.export fails with torch_compilable_check", - "right": "issue:44610 \u2014 OmDet-Turbo processor uses wrong image size", + "left": "issue:44062", + "right": "issue:44987", "accept": false, - "reason": "Different layers entirely: export/compilation versus processor configuration." + "reason": "Tokenizer construction bug vs model-loading regression; not the same underlying defect." }, { - "left": "issue:44062 \u2014 AddedToken gets duplicate special kwarg", - "right": "issue:44991 \u2014 transformers>=5.0.0 fails loading tokenizer for est-roberta", + "left": "issue:44265", + "right": "issue:44610", "accept": false, - "reason": "Both tokenizer-related, but one is an AddedToken constructor error and the other is a model-specific load failure." + "reason": "Export/compilation failure from torch_compilable_check is unrelated to OmDet processor image-size mismatch." }, { - "left": "issue:43322 \u2014 segmentation fault loading Llava Next", - "right": "issue:43540 \u2014 ValueError processing video inputs in Qwen3OmniMoe", + "left": "issue:43322", + "right": "issue:43540", "accept": false, - "reason": "Both are multimodal issues, but the failing models and failure modes are different." + "reason": "Different multimodal failures: Llava Next segfault vs Qwen3OmniMoe video-processing ValueError." }, { - "left": "issue:44849 \u2014 Qwen3.5 output_hidden_states bug", - "right": "issue:45325 \u2014 Qwen2.5-VL rope index scales still-image positions", + "left": "issue:42915", + "right": "issue:44589", "accept": false, - "reason": "Different models and separate bugs in hidden-state reporting vs position-id scaling." + "reason": "Different low-level failures: FP8 config handling vs missing Float8 storage type." }, { - "left": "issue:44877 \u2014 strict config prevents granite_speech loading", - "right": "issue:46661 \u2014 remote-code model_class lacks config_class", + "left": "issue:43632", + "right": "issue:46821", "accept": false, - "reason": "Both concern config handling, but one is strict config validation and the other is remote-code class metadata." + "reason": "Both are v5-era regressions, but one is the _is_hf_initialized flag and the other is a TimesFM window_size attribute error." }, { - "left": "issue:44987 \u2014 transformers>=5.1.0 fails loading physical-intelligence/fast", - "right": "issue:45081 \u2014 Mistral tokenizer regex patch crashes", - "accept": false, - "reason": "No shared root cause beyond both being loading regressions." - }, - { - "left": "issue:44488 \u2014 current version cannot load cjvt/sleng-bert", - "right": "issue:45081 \u2014 Mistral tokenizer regex patch crashes", + "left": "issue:44479", + "right": "issue:45325", "accept": false, - "reason": "Different tokenizer/model failures with no concrete overlap." + "reason": "Related Qwen VL family, but one is a video-input regression and the other is still-image position-id scaling." }, { - "left": "issue:44464 \u2014 chunked generation inconsistent with compiled forward", - "right": "issue:45987 \u2014 StaticCache.get_seq_length returns Tensor instead of int", + "left": "issue:43883", + "right": "issue:44991", "accept": false, - "reason": "Both touch generation internals, but the bugs are not the same." + "reason": "Both are loading/tokenizer-adjacent, but they are different models and different error modes." }, { - "left": "issue:43089 \u2014 generation overhead from many GPU syncs", - "right": "issue:46821 \u2014 TimesFM window_size AttributeError", + "left": "issue:43089", + "right": "issue:45987", "accept": false, - "reason": "Performance overhead versus an attribute error in a different model." + "reason": "Generation performance overhead is not the same as an incorrect cache-length return type." }, { - "left": "issue:44589 \u2014 cannot find Float8_e4m3fnStorage", - "right": "issue:46869 \u2014 YOLO model failing on multi-GPU auto device map", + "left": "issue:43531", + "right": "issue:44877", "accept": false, - "reason": "Unrelated runtime failures: storage-type lookup vs device-mapping assert." + "reason": "No shared bug mechanism; sliding-window behavior and strict config validation are unrelated." }, { - "left": "issue:43531 \u2014 sliding_window issue with Qwen3-MoE", - "right": "issue:44877 \u2014 strict config prevents granite_speech loading", + "left": "issue:36032", + "right": "issue:41720", "accept": false, - "reason": "Different models and different failure types; no shared underlying bug." + "reason": "Tokenizer method conflict and multi-GPU Qwen device-map crash are unrelated." } ] }, @@ -3801,8 +3750,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -3888,220 +3837,226 @@ "issue:43296|issue:46821", "issue:43611|issue:43646" ], - "prepared_review_unit_hash": "edb86ed1f6feaaf002d2574da44be372f8a3bdf572dddc2621343fa9c629226c", + "prepared_review_unit_hash": "0d7a09e13ff5d474683eff1edce4427373aeec342184cc8b1b483404b4f53841", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36958, - "estimated_input_tokens": 9240, - "estimated_eval_tokens": 18736 + "serialized_chars": 36961, + "estimated_input_tokens": 9241, + "estimated_eval_tokens": 18738 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a grab bag of unrelated bug reports. I only see two likely duplicate mini-clusters: Qwen3.5 save_pretrained checkpoint regressions, and AutoTokenizer/tokenizer.json loading regressions. Everything else is just loosely related by model family or subsystem.", - "confidence": 0.63, - "canonical_issue_reason": "Issue 44462 is the most reusable root-cause report here: it describes the generic AutoTokenizer/tokenizer.json loading bug in a model-agnostic way, and it can naturally absorb the Granite-specific symptom.", + "summary": "Mostly unrelated bug reports. The only clear duplicate-like pair is the Qwen3.5 save_pretrained regression, where the later report appears to be the same checkpoint/key corruption issue resurfacing.", + "confidence": 0.74, + "canonical_issue_reason": "issue:45216 is the clearest concrete regression and has the strongest duplicate-style overlap with issue:45357 about Qwen3.5 save_pretrained producing incorrect checkpoint contents.", "canonical_pr_reason": null, - "best_issue_reason": "Best anchor for triage because it states the underlying tokenizer-loading regression broadly rather than a single downstream symptom.", + "best_issue_reason": "issue:45216 is the best representative issue because it is specific, regression-focused, and has the clearest evidence of a same-bug follow-up in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Same Qwen3.5 save_pretrained regression; the later report narrows the symptom to incorrect visual encoder keys, but the underlying bug appears shared." + "reason": "Same Qwen3.5 save_pretrained regression; both report incorrect saved checkpoint/visual encoder keys in v5.x." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Different problems: torch version compatibility vs. noisy model loading output." + "reason": "Torch 2.10 compatibility failure vs noisy load logging; different bugs." }, { "left": "issue:44991", "right": "issue:46315", "accept": false, - "reason": "Both involve tokenizers in v5, but one is a loading failure and the other is a performance regression in convert_tokens_to_ids." + "reason": "Tokenizer loading failure vs slow convert_tokens_to_ids regression; related area but different concrete problems." }, { "left": "issue:40444", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen2.5-VL issues, but one is multi-image finetuning failure and the other is a rope/position_ids scaling bug." + "reason": "Both involve Qwen2.5-VL, but one is dataset/mult-image finetuning failure and the other is a rope/indexing bug." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "Different models and failure modes: PaddleOCR-VL load failure vs. Qwen3-MoE sliding_window behavior." + "reason": "PaddleOCR-VL load failure vs Qwen3-MoE sliding_window issue; unrelated." }, { "left": "issue:43716", "right": "issue:46899", "accept": false, - "reason": "Both mention multimodal dtype problems, but they affect different models and different code paths." + "reason": "Different multimodal dtype issues on different models and code paths; not the same bug." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "Separate v5 regressions: model loading with base_model_prefix vs. non-persistent buffer initialization." + "reason": "Both are v5 regressions during model loading, but one is base_model_prefix handling and the other is buffer initialization; not the same defect." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Different subsystems: fbank feature extraction vs. compiled flex_attention on newer torch." + "reason": "Audio feature extraction bug vs compiled flex_attention torch-version issue; unrelated." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "One is about noisy loading logs, the other about tokenizer loading failure; not the same bug." + "reason": "Generic load noise vs tokenizer loading failure; different symptoms and fixes." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but one is Granite token IDs and the other is CodeLlama leading-space round-trip behavior." + "reason": "Both are tokenizer regressions, but one is Granite token IDs and the other is CodeLlama round-trip spacing; not the same issue." }, { "left": "issue:42796", "right": "issue:43746", "accept": false, - "reason": "Unrelated issues: tokenizer BPE-marker handling vs. PEFT adapter checkpoint loading." + "reason": "Tokenizer BPE-marker handling vs PEFT/local-checkpoint loading for GraniteSpeech; unrelated." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Related only at a very high level; one is eval-mode checkpointing usage, the other is a missing capability flag on a trust_remote_code model." + "reason": "Eval-mode gradient checkpointing limitation vs missing supports_gradient_checkpointing flag; different code paths." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "Different regressions: _is_hf_initialized handling vs. mmap/OOM behavior on Strix Halo." + "reason": "HF initialization flag regression vs mmap/OOM workaround; unrelated." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Both are Qwen-VL-related, but multi-GPU training and multi-image finetuning are different problems." + "reason": "Qwen2-VL multi-GPU training vs Qwen2.5-VL multi-image dataset failure; both VL training, but not the same bug." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper runtime failure vs. missing Float8 storage type; no shared underlying code path is evident." + "reason": "Whisper run failure vs Float8 storage lookup error; different failures." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Completely different models and symptoms: high perplexity vs. processor.batch_decode skip_special_tokens handling." + "reason": "Perplexity regression on GPT-OSS vs Whisper batch_decode special-token handling; unrelated." }, { "left": "issue:42796", "right": "issue:46710", "accept": false, - "reason": "Tokenizer BPE-marker handling and accuracy regression after upgrade are not the same concrete bug." + "reason": "Tokenizer marker bug vs accuracy regression on a different model; not the same change." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "Different initialization failures: unexpected _is_hf_initialized argument vs. missing Float8 storage object." + "reason": "init_empty_weights/_is_hf_initialized argument issue vs missing Float8 storage object; different bugs." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multi-image finetuning failures." + "reason": "Progressive generation with inputs_embeds/past_key_values vs multi-image finetuning failure; unrelated." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "Different loading failures: Qwen2.5 GGUF/v5 compatibility vs. Llama4Config missing pad_token_id." + "reason": "Qwen2.5-GGUF loading failure vs Llama4Config pad_token_id attribute error; different models and failures." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment and Gemma4 FSDP2 shared_kv_states bugs are unrelated." + "reason": "ModelOutput key assignment bug vs Gemma4 FSDP2 shared_kv_states KeyError; unrelated." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to init_empty_weights/_is_hf_initialized errors." + "reason": "Qwen3-MoE sliding_window bug vs init_empty_weights argument bug; not the same problem." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "The first is a backend feature about updating post-processors; the second is a model-specific BOS/EOS addition bug." + "reason": "Runtime special-token post-processor update bug vs add_special_tokens BOS/EOS regression; similar area but different concrete failure." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Different model families and failure sites: Qwen3 auto device mapping vs. Mistral tokenizer regex patching." + "reason": "Auto device mapping cudaErrorAssert vs Mistral regex patch tokenizer crash; unrelated." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Both touch TensorFlow packaging/backends, but the dependency/install contexts and symptoms differ." + "reason": "Missing TensorFlow backend in torch-only install vs tensorboard-without-TF backend error; different packaging paths." }, { "left": "issue:43482", "right": "issue:45538", "accept": false, - "reason": "GGUF loading compatibility and CLIPTokenizer model_max_length are unrelated bugs." + "reason": "GGUF loading failure vs CLIPTokenizer max_length default bug; unrelated." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "Different regressions in different code paths: ModelOutput key handling vs. buffer initialization." + "reason": "ModelOutput key bookkeeping vs non-persistent buffer junk values; different internals." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Device-map/offload loading failure and Qwen3 MoE training memory inefficiency are different issues." + "reason": "device_map=auto dequantized-model offload failure vs Qwen3 MoE memory usage regression; not the same bug." }, { "left": "issue:44462", "right": "issue:45812", - "accept": true, - "reason": "Likely the same underlying AutoTokenizer/tokenizer.json regression; Granite wrong token IDs look like a concrete manifestation of the generic repository-loading bug." + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json vs wrong Granite token IDs; different tokenizer regressions." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "PaddleOCR-VL load failure vs TimesFM window_size AttributeError; unrelated." }, { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both are v5 model-init/loading regressions, but the titles indicate different failure modes and not one clear shared bug." + "reason": "Both are Transformers v5 loading regressions, but one is base_model_prefix handling and the other is custom model initialization; too broad to merge as one bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded and conservative. The summary sticks to the packet titles and only promotes two plausible duplicate clusters, which are both reasonably supported: Qwen3.5 save_pretrained regressions and AutoTokenizer/tokenizer.json regressions. The accepted soft pairs look defensible, and the rest of the verdicts err on the side of rejecting weakly related issues rather than overstating duplicates." + "feedback": "The summary is grounded in the packet and stays conservative. The accepted 45216\u219445357 pair is well supported by the titles, while the rejected pairs are mostly distinct bugs with appropriately cautious reasoning." }, "evaluator_used": true, "retried": false, @@ -4117,8 +4072,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -4203,221 +4158,224 @@ "issue:41720|issue:45823", "issue:43454|issue:43550" ], - "prepared_review_unit_hash": "0810c5943c592c9896496184e80d0db492fb738b6de80cac54ecdd6a95f1562d", + "prepared_review_unit_hash": "a6239947f8ec760c93c054f340644e850f3b1a16fd364098623e06130f1b1a93", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36923, - "estimated_input_tokens": 9231, - "estimated_eval_tokens": 18718 + "serialized_chars": 36926, + "estimated_input_tokens": 9232, + "estimated_eval_tokens": 18720 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a heterogeneous cluster of unrelated issues across models, tokenization, quantization, device mapping, and docs. None of the soft-edge pairs are convincing duplicates of the same underlying bug/change.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "The cluster is mostly heterogeneous, but two soft-related subthreads stand out: the SigLIP2 implementation/output regression pair and the Qwen2.5-VL visual position-id/rope-index regression pair. Those matches look like the closest counterparts in the packet, but they should be treated as likely related rather than confirmed duplicate identities.", + "confidence": 0.78, + "canonical_issue_reason": "No single canonical issue cleanly represents the whole cluster because the items span unrelated bugs. If a representative issue is needed, issue:43493 is the best closest counterpart for the SigLIP2 subthread because it describes the root implementation discrepancy behind the downstream bad outputs seen in 43994.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:43493 is the strongest issue anchor overall: it is concrete and root-cause oriented, and it is the best closest counterpart for the SigLIP2-related reports. issue:45325 is the other strong candidate for the Qwen2.5-VL position-id regression subthread, but it is a separate thread.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44509", "right": "issue:46909", "accept": false, - "reason": "Both are documentation issues, but they target different docs/examples and different models/tasks." + "reason": "Both are documentation bugs, but they concern different docs/examples and different failure modes; not the same underlying issue." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "One is about an install extra missing TensorFlow; the other is a PaddleOCR-VL load failure in vLLM." + "reason": "Unrelated problems: backend packaging vs a PaddleOCR-VL load failure." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both involve batch/padding behavior, but they affect different models and different failure modes." + "reason": "Both mention batching/padding, but they are different models and different bugs." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Library-internal regressions, but one is junk-filled non-persistent buffers and the other is ModelOutput key assignment." + "reason": "Different code paths: non-persistent buffer init vs ModelOutput key assignment." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen VL-related, but one is training with multiple images per prompt and the other is a video position-id regression." + "reason": "Same broad model family, but one is finetuning with multiple images and the other is a video position-id regression; not the same concrete bug." }, { "left": "issue:43818", "right": "issue:46402", "accept": false, - "reason": "Different models and code paths: Video-LLaVA architecture/weights vs Sam3Model text encoder loading." + "reason": "Different models and failures: video_tower architecture issue vs text-encoder weight loading." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Unrelated model families; one is Video-LLaVA internals, the other is Qwen3.5 save_pretrained key naming." + "reason": "Different models and different save/load bugs; only a broad visual-component theme overlaps." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Both touch multimodal indexing, but the models and concrete errors are different." + "reason": "Both involve position/index logic, but the model, modality, and failure are different." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export with torch_compilable_check is unrelated to dtype mismatches in bfloat16 inference." + "reason": "Different failures in different subsystems: export handling vs dtype mismatch in inference." }, { "left": "issue:43493", "right": "issue:43994", - "accept": false, - "reason": "Same model family, but one is a HF-vs-JAX implementation discrepancy while the other is a nonsensical AutoModel/pipeline result; not clearly the same bug." + "accept": true, + "reason": "Same SigLIP2 correctness thread: 43493 describes the HF-vs-JAX implementation discrepancy, and 43994 shows the downstream nonsensical outputs from that bug." }, { "left": "issue:45325", "right": "issue:45381", - "accept": false, - "reason": "Both are Qwen2.5-VL position-id bugs, but one is still-image scaling and the other is video vision_position_ids." + "accept": true, + "reason": "Likely the same Qwen2.5-VL visual position-id / rope-index bug affecting image/video inputs, though the packet only supports a soft duplicate read." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Both concern bad sequence-classification behavior, but one is tied to flash_attention_3 training and the other to num_labels=1/problem_type handling." + "reason": "Different training pathologies with different triggers; not the same bug." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Both mention lm_head/tied weights, but one is a config loading/tie failure and the other is serialization inconsistency by device." + "reason": "Both mention tied lm_head weights, but one is a model-specific tie failure and the other is a serialization inconsistency." }, { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "Completely different areas: tokenizer loading error message vs Apple Silicon allocator TypeError." + "reason": "Completely unrelated: tokenizer error messaging vs Apple Silicon allocator TypeError." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Different quantization/loading failures on different models; only loosely related by transformer version/quantization." + "reason": "Different quantization/load paths and different models; only a generic loading failure overlaps." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "Both are load-time failures, but the causes are distinct: unexpected init_empty_weights arg vs strict config rejection." + "reason": "Different root causes: init_empty_weights argument mismatch vs strict config loading." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen2.5-VL rope/position-id scaling." + "reason": "Checkpoint resume key mismatch is unrelated to visual position-id computation." }, { "left": "issue:42796", "right": "issue:45381", "accept": false, - "reason": "Tokenizer BPE-marker handling is unrelated to Qwen2.5-VL video position ids." + "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video position ids are different bugs." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to tokenization changes across versions." + "reason": "Quantized offloading vs version-dependent tokenization are unrelated." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Perplexity on gpt-oss-20b is unrelated to Whisper empty-transcription failure." + "reason": "Different models and symptoms: perplexity evaluation vs empty-transcription generation failure." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "Device mismatch with CPU offload is unrelated to torch.export failures around torch_compilable_check." + "reason": "Different runtime failures: CPU offload device mismatch vs export-time torch_compilable_check issues." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Both are regression/device issues, but one is Qwen video-input regression and the other is Gemma4 PLE device mismatch." + "reason": "Different models and different device-map problems; too broad to be the same bug." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio sampling-rate defaults and PEFT checkpoint loading are different speech-model bugs." + "reason": "Audio chat-template sampling-rate defaulting and GraniteSpeech checkpoint loading are unrelated." }, { "left": "issue:43828", "right": "issue:46821", "accept": false, - "reason": "Different model/runtime failures: autocast dtype mismatch vs TimesFM window_size AttributeError." + "reason": "Different failures: autocast dtype mismatch vs missing window_size attribute." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 auto device mapping CUDA asserts." + "reason": "Different model behaviors and different failure modes; no shared underlying bug." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice preset and progressive generation with cache/embeds are unrelated bugs." + "reason": "BarkProcessor voice preset and progressive generation with inputs_embeds/past_key_values are unrelated." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "A documentation example error for SigLIP2 is unrelated to Qwen2.5-VL finetuning with multiple images." + "reason": "Both are multimodal/model-usage issues, but they are about different models and distinct bugs." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer crashes on NER padding/truncation are unrelated to export failures from torch_compilable_check." + "reason": "Tokenizer padding/truncation crashes are unrelated to torch.export exportability." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Both involve device_map=\"auto\", but they are different models with different concrete failure modes." + "reason": "Different models and different device-mismatch mechanisms; not the same bug." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "AyaVision weight-tying/generation corruption is unrelated to Bamba torch.compile SDPA failures." + "reason": "Tied-weight generation bug and torch.compile SDPA failure are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level read is mostly grounded, but a few phrases overstate the evidence. In particular, calling issue:43994 a \"true-duplicate counterpart\" of issue:43493 is stronger than the packet supports, since the available material only shows a similar SigLIP2 root-cause/symptom relationship, not an explicit duplicate link. Likewise, the 45325\u219445381 match is plausible but should be described more cautiously as a likely or soft duplicate, not as confirmed same-bug identity. Please soften those claims and keep the anchor language to \"closest counterpart\" / \"likely related\" rather than definitive duplicate wording." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -4429,8 +4387,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -4515,214 +4473,214 @@ "issue:26413|issue:28530", "issue:43873|issue:44756" ], - "prepared_review_unit_hash": "88906229a1dce95a01f7e632f346770424310edd81f596619e9fadb24d9bddf1", + "prepared_review_unit_hash": "ee4f1fb3c5d26402a76cf03d9b138ca90310095bef26203c36aae59ba332ef40", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36631, - "estimated_input_tokens": 9158, - "estimated_eval_tokens": 18572 + "serialized_chars": 36636, + "estimated_input_tokens": 9159, + "estimated_eval_tokens": 18574 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose group of Transformers bug reports around loading/offloading, quantization, tokenizer regressions, model-specific failures, and a few training/runtime issues. The soft pairs are semantically related in broad subsystem terms, but they are not the same underlying bug and should stay separate.", - "confidence": 0.81, - "canonical_issue_reason": "issue:43873 is the broadest, most central report in the set: offloading plus quantization/device-placement problems, with multiple related references and clear overlap with several nearby loading bugs.", + "summary": "These are mostly distinct transformer/model bugs with only superficial overlap (often the same broad area like offloading, tokenization, or saving/loading). I found no soft pair that is clearly the same underlying bug/change, so all pairwise duplicate links should be rejected.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43873 is the broadest and most centrally referenced issue in the set, with a general offloading/quantization regression theme and the highest inbound-reference count among the candidates.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the best representative issue for the cluster because it captures the shared theme of loading/offload/quantization regressions without overfitting to one model-specific failure.", + "best_issue_reason": "issue:43873 is the best single issue candidate because it is broad, well-discussed, and likely to capture related offloading/quantization regressions better than the narrower model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "Both involve model loading/offload behavior, but one is a CPU-offload device mismatch for bark-small and the other is DAC.from_latents vs forward mismatch due to missing STE; different bugs and code paths." + "reason": "Different models and different failures: CPU-offload device mismatch vs DAC latent/forward mismatch." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Both mention Qwen3 and configuration/loading concerns, but one is about MOE weights not converting under accelerate+deepspeed while the other is a missing gradient_checkpointing flag on a trust_remote_code model; not the same issue." + "reason": "Qwen3 MoE loading/DeepSpeed issue versus a trust_remote_code gradient-checkpointing flag omission." }, { "left": "issue:42796", "right": "issue:46209", "accept": false, - "reason": "Tokenizer BPE-marker handling for Ministral-3-8B-Instruct is unrelated to FP8 inference failure on B200 hardware." + "reason": "Tokenizer BPE-marker handling bug is unrelated to FP8 inference failure on B200." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is about add_special_tokens not adding BOS/EOS for a specific tokenizer, while the other is wrong token IDs across multiple models; different underlying failures." + "reason": "Both are tokenizer regressions, but one is a specific BOS/EOS special-token bug and the other is a broad wrong-ID regression across several models." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "Both are shape/consistency bugs in multimodal/audio-image processors, but one is a padding-mask batching issue in MIMI Encoder and the other is an OmDet-Turbo image-size mismatch; not duplicates." + "reason": "Both involve batched inputs, but they concern different components and different concrete bugs." }, { "left": "issue:45823", "right": "issue:46821", "accept": false, - "reason": "Both concern windowing/offloading-style API behavior, but one is Gemma4 device mismatch with auto device map and the other is a TimesFM window_size AttributeError; separate models and failures." + "reason": "Device-mapping forward crash versus a TimesFM window_size AttributeError; unrelated code paths." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "AyaVision lm_head weight tying causing garbage generation is unrelated to SwitchTransformers/TimmWrapperModel bfloat16 dtype mismatches." + "reason": "lm_head weight tying/serialization problem is not the same as a bfloat16 dtype mismatch." }, { "left": "issue:46305", "right": "issue:46315", "accept": false, - "reason": "Both are regression-style bugs involving window/token processing, but one is an off-by-one in MoonshineStreaming mask logic and the other is a tokenizer performance regression; not the same code-path bug." + "reason": "Sliding-window mask off-by-one is unrelated to tokenizer conversion performance." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "High perplexity on a specific model/dataset and ineffective gradient checkpointing with PEFT LoRA are distinct issues." + "reason": "High perplexity on a model is not the same bug as PEFT LoRA gradient-checkpointing ineffectiveness." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "A generic 3d_parallel.py runtime failure and a Qwen3MoE fine-grained FP8 config failure are unrelated." + "reason": "3d_parallel.py runtime failure is unrelated to Qwen3 MoE fine-grained FP8 loading." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Both involve processor/input handling, but one is batched-vs-single output inconsistency from missing padding masks and the other is non-default-resolution Perceiver classification failure despite interpolate_pos_encoding=True; different model paths." + "reason": "Different vision/input bugs: padding-mask batch mismatch versus incorrect resolution handling in Perceiver." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "A missing pad_token_idx in StableLmConfig and failure to unwrap text_config for composite Qwen multimodal configs are different configuration bugs." + "reason": "Missing pad_token_idx in one config is unrelated to unwrapping text_config for composite Qwen multimodal configs." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Both relate to offloading/quantization, but one is dequantized model loading with device_map=auto on gpu+cpu offload and the other is a broader offloading-not-working report; related theme, not the same defect." + "reason": "Both mention offloading, but one is dequantized-model loading and the other is broader quantization/offload behavior; not the same concrete bug." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "Different model families and failures: weight tying in AyaVision versus packed-weight unpacking in BitNet loading." + "reason": "Different serialization/loading issues: tied lm_head weights versus packed BitNet weights not being unpacked." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "FSDP_CPU_RAM_EFFICIENT_LOADING breakage and non-persistent buffers being filled with junk are separate runtime/regression issues." + "reason": "FSDP CPU RAM-efficient loading failure is unrelated to non-persistent buffers being filled with junk." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "glm-4v-9b loading failure and cjvt/sleng-bert loading failure are different model-specific load problems." + "reason": "Two model-loading failures, but for different models and different root causes." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "Llava Next segmentation fault and sinq quantized model save/load breakage are not the same underlying bug." + "reason": "Segfault when loading Llava Next is unrelated to sinq quantization save/load corruption." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert on A800 is a device-mapping/runtime assertion issue, while Llava Next is a load-time segfault; different code paths." + "reason": "Both are severe load-time failures, but for different models and failure modes (cudaErrorAssert vs segfault)." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "CustomTrainer compute_loss signature regression and inability to run 3d_parallel.py are unrelated training/runtime issues." + "reason": "Trainer signature change bug is unrelated to a parallel script runtime failure." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Both mention Qwen3 and flash/device execution, but one is an auto device-mapping crash and the other is a flash-attention generation bug; not the same concrete defect." + "reason": "Same model family, but different concrete bugs: device-mapping assert versus flash-attention generation failure." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "Int4 quantization OOM due to increased CUDA reserved memory is not the same as T5 silently using apex.FusedRMSNorm with a memory leak." + "reason": "Memory pressure under int4 quantization is not the same as a T5 Apex RMSNorm memory leak." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Unexpected compute_loss keyword argument and Qwen3 MOE fine-grained FP8 loading failure are unrelated." + "reason": "Unexpected compute_loss keyword argument is unrelated to Qwen3 MoE FP8 loading." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config blocking granite_speech and tokenizer loading failure for est-roberta are different loading/configuration bugs." + "reason": "Strict config loading failure and tokenizer-loading regression are different code paths." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "High perplexity on one model/dataset is not the same as version-dependent tokenization changes." + "reason": "Perplexity regression is unrelated to tokenization changing across versions." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Both are serialization/tied-weight regressions in v5, but one is lm_head.weight serialization depending on device and the other is incorrect visual encoder keys in Qwen3.5 save_pretrained; different artifacts and fixes." + "reason": "Both are save_pretrained regressions, but they affect different tensors/keys and likely different fixes." }, { "left": "issue:43450", "right": "issue:45701", "accept": false, - "reason": "Video processor batched shape bug and version-dependent tokenization changes are separate regressions." + "reason": "Incorrect video batch shape is unrelated to tokenization changes." }, { "left": "issue:43065", "right": "issue:46402", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder and missing text encoder weight loading in Sam3Model are related by model family but not the same bug." + "reason": "Sam3 pixel-decoder dummy Conv2d is unrelated to Sam3 text-encoder weight loading." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and a generic 3d_parallel.py failure are different runtime problems." + "reason": "Different systems and failure modes: auto device-mapping assert versus 3d_parallel script failure." }, { "left": "issue:26413", "right": "issue:28530", "accept": false, - "reason": "Checkpoint resume epoch_iterator failure and early-stopping metric lookup failure are unrelated training-loop issues." + "reason": "Checkpoint resume bug is unrelated to early-stopping metric resolution." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Quantization/offloading instability is different from disabling mmap on Strix Halo to avoid OOM; the mechanism and fix differ." + "reason": "Quantization/offloading regression is unrelated to disabling mmap to avoid Strix Halo OOM." } ] }, @@ -4825,202 +4783,202 @@ "issue:44671|issue:44743", "issue:45632|issue:45684" ], - "prepared_review_unit_hash": "30fc4eb86a7bc25fc43469394d1b4649d6bce401ef9c6df92e71d68851cff178", + "prepared_review_unit_hash": "efff030601e9dd4dcb5469a3b985c8c31ff469f2ba2f3d4ecc012d8c99e419d1", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36298, - "estimated_input_tokens": 9075, - "estimated_eval_tokens": 18406 + "serialized_chars": 36301, + "estimated_input_tokens": 9076, + "estimated_eval_tokens": 18408 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No duplicate-level cluster emerges here: the items span many unrelated bugs across trainers, tokenizers, generation, multimodal serving, quantization, and device/offload paths. All soft-edge candidates are rejected.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "These are mostly unrelated bug reports across training, tokenizers, multimodal processors, generation cache logic, and model loading. The soft overlaps are only superficial; none look like the same underlying bug or change.", + "confidence": 0.89, + "canonical_issue_reason": "issue:46858 is the most concrete, self-contained bug report in the set and is closest to the cache/generation cluster, but there is no true duplicate hub here.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:46858 has the clearest symptom, reproduction, and code-path specificity among these unrelated reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Different model families and failure modes: tokenizer/padding crash vs bfloat16 dtype mismatch." + "reason": "Different subsystems and failures: tokenizer/NER padding crash vs dtype mismatch in inference." }, { "left": "issue:46612", "right": "issue:46858", "accept": false, - "reason": "Both involve cache logic, but one is beam-search reorder skipping and the other is static-cache slice arithmetic on non-CUDA backends." + "reason": "Both involve generation cache code, but the bugs are different: skipped beam cache reorder vs static-cache seq_length arithmetic on non-CUDA backends." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Unrelated bugs: BLIP2 dtype propagation vs Trainer resume/max_steps calculation." + "reason": "Unrelated issues: BLIP2 dtype loading vs Trainer resume max_steps calculation." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "Different code paths: torch.compile+SDPA in Bamba vs packed-weight unpacking during accelerate load." + "reason": "Different concrete failures: SDPA with torch.compile vs packed weights not unpacked during accelerate loading." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "Both mention tokenization regressions, but one is a processor/tokenizer API break and the other is a version-dependent tokenization change." + "reason": "Both are regressions around loading/tokenization behavior, but one is processor.tokenizer access and the other is version-dependent tokenization differences." }, { "left": "issue:45685", "right": "issue:46821", "accept": false, - "reason": "Unrelated: MoE MPS histogram kernel error vs TimesFM window_size attribute error." + "reason": "No shared bug: MPS histogram support in MoE vs TimesFM window_size attribute access." }, { "left": "issue:43646", "right": "issue:46665", "accept": false, - "reason": "Shared custom-model loading theme, but different defects: initialization breakage vs from_pretrained behavior inside Python classes." + "reason": "Both mention custom loading flows, but one is Transformers 5 custom model init breakage and the other is from_pretrained inside Python classes." }, { "left": "issue:44451", "right": "issue:47020", "accept": false, - "reason": "Different tokenizer-loading bugs affecting different models and root causes." + "reason": "Tokenizer loading failures, but for different root causes and model families; not the same underlying issue." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "Different targets: GLM-4V loading failure vs tokenizer load failure for est-roberta." + "reason": "Different failures: glm-4v-9b load error vs tokenizer loading for EMBEDDIA/est-roberta." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "Unrelated: BeitImageProcessorFast label reduction vs missing all_tied_weights_keys on Molmo." + "reason": "Completely different bugs: BEiT reduce_labels output vs missing all_tied_weights_keys on MolmoForCausalLM." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Both are vision-input issues, but one is tokenizer crashes and the other is processor output size mismatching the model." + "reason": "Both are preprocessing/processor issues, but one is text/tokenization padding and the other is image size mismatch." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Different bugs: compile/SDPA failure vs CPU-offload device mismatch." + "reason": "Different code paths: torch.compile SDPA on Bamba vs CPU offload device mismatch in Bark." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "Different failures: GenerationMixin import error vs a 3d_parallel script runtime issue." + "reason": "No meaningful overlap: GenerationMixin import error vs a 3d_parallel.py runtime failure." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "Different model-specific save/load bugs: tied lm_head weights vs wrong visual encoder keys." + "reason": "Different model-save/load regressions: tied lm_head weights vs incorrect visual encoder keys." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Both involve loading/quantization, but one is FineGrainedFP8 on Qwen3Moe and the other is device_map/offload loading." + "reason": "Both are model-loading/runtime failures, but the causes differ: FP8 config failure vs dequantized model offload failure." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Similar image-size theme, but different models and mechanics: processor resolution mismatch vs interpolation failure." + "reason": "Different model families and failure modes: OmDet-Turbo image-size mismatch vs Perceiver positional encoding failure." }, { "left": "issue:46821", "right": "issue:46869", "accept": false, - "reason": "Different areas: TimesFM window_size attribute access vs YOLO multi-GPU device_map loading." + "reason": "Unrelated: TimesFM window_size AttributeError vs YOLO multi-GPU device-map loading." }, { "left": "issue:44734", "right": "issue:46858", "accept": false, - "reason": "Both touch cache arithmetic, but one is serve KV-cache continuation indexing and the other is static-cache seq_length on non-CUDA compiled prefill." + "reason": "Both touch cache behavior, but one is KV-cache continuation indexing in serve and the other is seq_length device arithmetic in static cache." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Both are serve/multimodal crashes, but one is string-to() misuse and the other is a missing processor tokenizer attribute." + "reason": "Both are serve/multimodal crashes, but the underlying bugs differ: string-to tensor conversion path vs Gemma4Processor missing _tokenizer." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Different Gemma issues: variable-name collision in forward vs device mismatch under auto device mapping." + "reason": "Same broad Gemma area, but different bugs: variable name collision in Gemma3n vs device mismatch in Gemma4 PLE." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Different parallelism bugs: auto device mapping assert vs tensor-parallel failure in GPT-OSS MoE." + "reason": "Different model-specific runtime issues: Qwen3 auto device-map assertion vs GPT-OSS MoE tensor-parallel behavior." }, { "left": "issue:28530", "right": "issue:33666", "accept": false, - "reason": "Completely different: early-stopping metric config vs Qwen2-VL multi-GPU training." + "reason": "Training early-stopping configuration bug vs Qwen2-VL multi-GPU training request." }, { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Different model/config bugs: sliding_window behavior in Qwen3-MoE vs missing pad_token_idx in StableLmConfig." + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs StableLmConfig missing pad_token_idx." }, { "left": "issue:44464", "right": "issue:46858", "accept": false, - "reason": "Both involve generation/compile, but one is inconsistent chunked outputs and the other is static-cache slice arithmetic from a device tensor." + "reason": "Both involve compiled generation, but one is inconsistent chunked outputs and the other is static-cache sequence-length arithmetic." }, { "left": "issue:42915", "right": "issue:46291", "accept": false, - "reason": "Different subsystems: MoE FP8 loading vs pipelines import failure due to kernels version/revision requirements." + "reason": "Completely unrelated: FP8 model failure vs kernels package import/version requirement." }, { "left": "issue:36331", "right": "issue:41720", "accept": false, - "reason": "Trainer compute_loss signature mismatch is unrelated to the A800 cudaErrorAssert in Qwen3 auto device mapping." + "reason": "Different layers and symptoms: Trainer API change in compute_loss vs a CUDA assert in Qwen3 auto device mapping." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "Different defects: CamemBERT masked-LM regression vs Qwen3.5 recurrent state reset with cache." + "reason": "Distinct model behaviors: CamemBERT masked-LM predictions vs Qwen3.5 recurrent state reset with cache." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "Both are custom-model file handling issues, but one is cache-path collision and the other is permissions propagation on save_pretrained." + "reason": "Different save_pretrained problems: cache-path collision vs read-only permission propagation." } ] }, @@ -5061,6 +5019,7 @@ "issue:43388", "issue:43425", "issue:43452", + "issue:43526", "issue:43531", "issue:43540", "issue:43550", @@ -5070,17 +5029,16 @@ "issue:43746", "issue:43782", "issue:43873", + "issue:43881", "issue:44186", "issue:44361", "issue:44387", - "issue:44464", "issue:44492", "issue:44509", "issue:44534", - "issue:44857", + "issue:44589", "issue:44913", "issue:45059", - "issue:45072", "issue:45406", "issue:45464", "issue:45701", @@ -5089,13 +5047,11 @@ "issue:45859", "issue:46665", "issue:46821", - "issue:47020", - "issue:47050" + "issue:46858", + "issue:47020" ], "soft_pairs": [ - "issue:44857|issue:47050", "issue:44186|issue:44361", - "issue:44464|issue:47050", "issue:43425|issue:46821", "issue:43316|issue:44913", "issue:28530|issue:30064", @@ -5105,7 +5061,6 @@ "issue:29159|issue:36032", "issue:30064|issue:34634", "issue:43873|issue:45704", - "issue:45859|issue:47050", "issue:44387|issue:45701", "issue:44534|issue:46665", "issue:43531|issue:45759", @@ -5115,209 +5070,219 @@ "issue:41720|issue:43746", "issue:43606|issue:43720", "issue:42490|issue:43329", - "issue:45072|issue:47050", "issue:45406|issue:45464", "issue:43388|issue:45059", "issue:34567|issue:35707", "issue:43531|issue:46821", - "issue:43606|issue:43653" + "issue:43606|issue:43653", + "issue:43526|issue:43881", + "issue:34634|issue:41720", + "issue:43540|issue:46858", + "issue:43531|issue:44589", + "issue:45759|issue:45859" ], - "prepared_review_unit_hash": "a6b6d7ea88a416407d2b8353400f8b5092a6026326142d0d94b9e28335073e07", + "prepared_review_unit_hash": "02c50ccfef6a82d78ac5b8adbceb0bdd4b5dadd32fa0f10cfc4e89a20c771014", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35509, - "estimated_input_tokens": 8878, - "estimated_eval_tokens": 18012 + "soft_pair_count": 29, + "serialized_chars": 35785, + "estimated_input_tokens": 8947, + "estimated_eval_tokens": 18150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "None of the soft-edge pairs look like true duplicates; they mostly describe different models, different subsystems, or different failure modes. If one issue has to serve as the best canonical representative, issue 47050 is the most concrete single-code-path bug report.", - "confidence": 0.84, - "canonical_issue_reason": "issue:47050 is the strongest canonical representative here because it describes a specific, reproducible dtype-mismatch failure in one code path (CanineModel fp16 ONNX export on CUDA) rather than a broad or multi-root report.", + "summary": "This cluster is a loose collection of unrelated Transformers issues; the soft-similarity pairs share vocabulary but not the same underlying bug, so none should be merged as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:47050 is the most actionable and narrowly scoped report in the set, making it the best anchor for triage among unrelated issues.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44857", - "right": "issue:47050", - "accept": false, - "reason": "Both are CUDA/dtype failures, but they affect different models and different pipelines (AMP loss vs ONNX export attention)." - }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Different tokenizers and different failure modes; not the same underlying bug." - }, - { - "left": "issue:44464", - "right": "issue:47050", - "accept": false, - "reason": "Chunked generation/compiled forward is unrelated to ONNX export dtype mismatch." + "reason": "Both are tokenizer/task crashes, but they affect different models and failure modes (LayoutLMv2 NER/padding vs MLuke task AttributeError)." }, { "left": "issue:43425", "right": "issue:46821", "accept": false, - "reason": "Torch version compatibility and a TimesFM window_size AttributeError are unrelated bugs." + "reason": "Unrelated problems: Torch 2.10 compatibility vs TimesFM window_size AttributeError." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Both are config/API inconsistencies, but they concern different models and different fields." + "reason": "Both mention config API behavior, but one is Gemma3TextConfig parity and the other is GPTNeoX rotary_pct reload persistence." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Early stopping metric selection is unrelated to image processor segmentation-map handling." + "reason": "Early-stopping metric selection and image processor void segmentation maps are different subsystems and bugs." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Different models and different failure classes (FP8 config error vs segmentation fault on load)." + "reason": "Different model-loading failures for different models; no shared concrete code path." }, { "left": "issue:33357", "right": "issue:35707", "accept": false, - "reason": "A MacOS CLIP bus error is not the same as progressive generation with inputs_embeds/past_key_values." + "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are unrelated." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Both are Qwen multimodal issues, but they affect different model variants and different stages (video processing vs from_pretrained weight loading)." + "reason": "Different Qwen model families and different from_pretrained/video-loading errors." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Inconsistent slow/fast tokenizer behavior is not the same as the T5 add_special_tokens constructor conflict." + "reason": "Tokenizer inconsistencies are too broad here; one is slow-vs-fast parity, the other is a T5 add_special_tokens name conflict." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Void segmentation map processing and BarkProcessor voice_preset handling are unrelated." + "reason": "Image segmentation-map processing and Bark voice_preset handling are unrelated processor bugs." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantization/offloading behavior and an apex RMSNorm memory leak are separate bugs." - }, - { - "left": "issue:45859", - "right": "issue:47050", - "accept": false, - "reason": "CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding is unrelated to a CanineModel ONNX dtype mismatch." + "reason": "Quantization/offloading behavior and T5 apex FusedRMSNorm memory leak are different problems." }, { "left": "issue:44387", "right": "issue:45701", "accept": false, - "reason": "Memory growth under int4 quantization is not the same as version-dependent tokenization changes." + "reason": "Int4 CUDA OOM and tokenization changes across versions are not the same issue." }, { "left": "issue:44534", "right": "issue:46665", "accept": false, - "reason": "Non-persistent buffer junk in v5 and from_pretrained behavior inside Python classes are different issues." + "reason": "Non-persistent buffer initialization junk and from_pretrained-in-python-class behavior are unrelated." }, { "left": "issue:43531", "right": "issue:45759", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE and text_config unwrapping in composite Qwen multimodal configs are different code paths." + "reason": "Both involve Qwen configs, but one is sliding_window on Qwen3-MoE and the other is text_config unwrapping for multimodal configs." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "torch.compile/SDPA failure is unrelated to tokenizer special-token registration." + "reason": "Different components and failures: torch.compile/SDPA vs tokenizer special-token registration." }, { "left": "issue:43452", "right": "issue:47020", "accept": false, - "reason": "Both involve loading/tokenizer init, but the root causes are different: gguf_file handling vs Unigram expecting a sequence." + "reason": "Both are loading/tokenizer-related, but one is GGUF auto-class loading and the other is XLMRobertaTokenizer passing the wrong type to Unigram." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A cache-strategy typo and outdated docs about removed pipeline tasks are not the same bug." + "reason": "A cache-strategy typo and deprecated docs for removed pipeline tasks are not the same change or bug." }, { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Auto device map cudaErrorAssert and PEFT-adapter checkpoint loading are unrelated." + "reason": "Different models and loading paths: Qwen3 auto device mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "Device mismatch during CPU offload and unpacking packed weights during accelerate loading are different failure mechanisms." + "reason": "Both involve accelerate/offload-like loading, but the models and root causes differ (device mismatch vs packed weight unpacking)." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "A trailing-slash local path import warning is unrelated to undefined video-token helper functions." - }, - { - "left": "issue:45072", - "right": "issue:47050", - "accept": false, - "reason": "bfloat16 inference dtype mismatches across two models are not the same as a CanineModel fp16 ONNX export bug." + "reason": "Local-path import warning and multimodal token counting video-branch undefined variables are unrelated." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat failures are different issues." + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat/completions failures do not share a concrete bug." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric gathering label truncation and SAM3 text/bounding-box behavior are unrelated." + "reason": "Metric gathering on tuple labels and SAM3 text/bounding-box behavior are different code paths." }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "TrainerState token counting and progressive generation with cache embeddings are different bugs." + "reason": "Trainer token-count state tracking and progressive generation are unrelated." }, { "left": "issue:43531", "right": "issue:46821", "accept": false, - "reason": "Both mention window/sliding_window concepts, but they involve different models and different failure surfaces." + "reason": "Different models and different bugs; no shared underlying fix." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Bark CPU offload device mismatch and BigBirdTokenizer special-token registration are unrelated." + "reason": "CPU offload device mismatch and BigBirdTokenizer special-token decode behavior are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and glm-4v-9b loading failure are unrelated." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset and Qwen3 auto device mapping are different subsystems and errors." + }, + { + "left": "issue:43540", + "right": "issue:46858", + "accept": false, + "reason": "Qwen3OmniMoe video-input validation and static-cache slice arithmetic on non-CUDA backends are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and Float8 storage lookup failure are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Different Qwen multimodal config/model issues: from_config text_config unwrapping vs CPU-offload rotary embedding." } ] }, @@ -5336,34 +5301,34 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ - "issue:34634", + "issue:30064", + "issue:36010", "issue:39692", "issue:41720", "issue:42886", "issue:43381", "issue:43421", - "issue:43526", - "issue:43531", - "issue:43540", + "issue:43493", "issue:43644", + "issue:43653", "issue:43746", + "issue:43782", "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:44589", "issue:44610", "issue:44625", "issue:44671", @@ -5376,10 +5341,10 @@ "issue:45042", "issue:45072", "issue:45357", + "issue:45381", "issue:45479", "issue:45704", "issue:45759", - "issue:45859", "issue:45923", "issue:46032", "issue:46050", @@ -5388,14 +5353,9 @@ "issue:46655", "issue:46665", "issue:46693", - "issue:46858" + "issue:47050" ], "soft_pairs": [ - "issue:43526|issue:43881", - "issue:34634|issue:41720", - "issue:43540|issue:46858", - "issue:43531|issue:44589", - "issue:45759|issue:45859", "issue:45357|issue:46402", "issue:44451|issue:44991", "issue:44077|issue:44625", @@ -5419,219 +5379,224 @@ "issue:45759|issue:45923", "issue:43746|issue:43873", "issue:43994|issue:44671", - "issue:46032|issue:46463" + "issue:46032|issue:46463", + "issue:39692|issue:43493", + "issue:43421|issue:43653", + "issue:30064|issue:36010", + "issue:43782|issue:45381", + "issue:44464|issue:47050" ], - "prepared_review_unit_hash": "f19daa392fd7b623f093934f7dbfd296b34be7e813e576691ab5b29bdc95e5d4", + "prepared_review_unit_hash": "9497bbcd809b970084854bfed706d46e9e1e1d20476042f3bb6e810b4ccee9ab", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36564, - "estimated_input_tokens": 9141, - "estimated_eval_tokens": 18538 + "serialized_chars": 36683, + "estimated_input_tokens": 9171, + "estimated_eval_tokens": 18598 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous cluster with two real duplicate pairs: the offline HF_HUB_OFFLINE tokenizer regression (42886 \u2194 44843) and the Qwen2_5_VLProcessor batched `padding=False` crash (44514 \u2194 44545). The rest look like separate bugs in different models/code paths.", - "confidence": 0.77, - "canonical_issue_reason": "Issue 44843 is the strongest representative for the offline tokenizer regression because it states the concrete failure and likely root cause (`model_info()` being called unconditionally), making it more actionable than the symptom-only report.", + "summary": "This cluster is mostly heterogeneous and contains only a couple of true duplicates. The strongest matches are the offline tokenizer/cache failure pair (42886 \u2194 44843) and the batched Qwen2_5_VLProcessor crash pair (44514 \u2194 44545).", + "confidence": 0.96, + "canonical_issue_reason": "The cluster does not have one unifying issue, but issue 44545 is the cleanest canonical anchor among the true duplicates because it is an exact title match with 44514 and describes a focused, reproducible processor crash.", "canonical_pr_reason": null, - "best_issue_reason": "44843 is the best cluster anchor because it is the clearest root-cause report for the offline tokenizer bug; 42886 is the closely related symptom report, but not a broader duplicate hub than the other pair.", + "best_issue_reason": "Issue 44545 is the best issue-level representative because it forms an exact duplicate pair with 44514; the rest of the items are largely unrelated bug reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43526", - "right": "issue:43881", - "accept": false, - "reason": "Different bugs: BEiT `reduce_labels` output shape vs a GLM-4V model loading failure." - }, - { - "left": "issue:34634", - "right": "issue:41720", - "accept": false, - "reason": "Unrelated failures: BarkProcessor `voice_preset` handling vs Qwen3 auto device mapping CUDA assert." - }, - { - "left": "issue:43540", - "right": "issue:46858", - "accept": false, - "reason": "Different code paths: Qwen3OmniMoe video-input processing vs cache length tensor arithmetic in `_prefill`." - }, - { - "left": "issue:43531", - "right": "issue:44589", - "accept": false, - "reason": "Different issues: Qwen3-MoE `sliding_window` behavior vs a Float8 storage lookup/type error." - }, - { - "left": "issue:45759", - "right": "issue:45859", - "accept": false, - "reason": "Different bugs: `from_config` unwrapping of composite Qwen configs vs CPU-offload rotary embedding incompatibility." - }, { "left": "issue:45357", "right": "issue:46402", "accept": false, - "reason": "Both are Qwen-related loading issues, but one is about wrong visual encoder keys on save and the other about missing text encoder weights on load." + "reason": "Both are model-loading regressions, but they affect different models and different failure modes: visual encoder keys vs missing text encoder weights." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading failures, but they affect different models and are likely separate root causes." + "reason": "Both are tokenizer/load failures, but they target different models and likely different code paths; not the same underlying bug." }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "Different model/config problems: optional `post_init` handling vs `num_labels` propagation." + "reason": "These are unrelated: one is about patchtsmixer post_init policy, the other about Qwen3.5 num_labels propagation." }, { "left": "issue:43644", "right": "issue:46665", "accept": false, - "reason": "Different failures: non-persistent buffers being filled with junk vs `from_pretrained` behavior inside Python classes." + "reason": "Different problems: non-persistent buffer initialization junk vs inconsistent from_pretrained behavior inside Python classes." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Separate bugs: loading `physical-intelligence/fast` vs PIL image processors wrongly requiring `torchvision`." + "reason": "One is a model load failure for a specific repo; the other is an image-processor dependency regression requiring torchvision." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "Different root causes: `AddedToken(..., special=...)` argument duplication vs ScandiBERT load failure." + "reason": "AddedToken argument duplication is a tokenizer construction bug, while the other is a model-specific load failure." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Different bugs: Perceiver resolution/interpolation handling vs bf16 dtype mismatches in inference." + "reason": "Perceiver interpolation at non-default resolution and bfloat16 dtype mismatches are unrelated code-path failures." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Both involve classification config/labels, but one is propagation of `num_labels` in Qwen3.5 and the other is a generic zero-loss `problem_type` bug." + "reason": "Both touch classification config, but one is missing num_labels propagation and the other is a degenerate loss bug." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Different training behaviors: gradient checkpointing in eval mode vs a ZeRO-3 gradient growth regression." + "reason": "Gradient checkpointing in eval mode and ZeRO-3 gradient explosion are different issues with different causes." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Both mention processor/model mismatches, but they are different models and different failure modes." + "reason": "A broken documentation example for SigLIP2 is unrelated to OmDet-Turbo producing the wrong processor image size." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "Unrelated Qwen3.5 warning/config issue vs incorrect visual encoder keys on save." + "reason": "A tie_word_embeddings warning in fine-tuning is unrelated to incorrect visual encoder keys saved by Qwen3.5." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "Different persistence/loading bugs: PEFT adapters failing from local checkpoints vs quantized sinq models breaking after save/load." + "reason": "Different load/save bugs: PEFT adapters not loading from local checkpoints vs quantized sinq models breaking after save/load." }, { "left": "issue:41720", "right": "issue:45704", "accept": false, - "reason": "Completely different problems: CUDA assert in Qwen3 auto device mapping vs T5/apex RMSNorm memory leak." + "reason": "Qwen3 auto device mapping cuda assert and T5 apex FusedRMSNorm memory leak are unrelated." }, { "left": "issue:46655", "right": "issue:46693", "accept": false, - "reason": "Different regressions: SmolVLM video padding calculation vs FlashAttention2 long-context performance regression." + "reason": "Incorrect video padding in SmolVLM is unrelated to a FA2 long-context performance regression." }, { "left": "issue:43381", "right": "issue:45704", "accept": false, - "reason": "Different issues: checkpointing/eval-mode restriction vs T5 apex memory leak." + "reason": "Gradient checkpointing eval-mode restrictions have nothing to do with T5's apex memory leak." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Different symptoms and code paths: Qwen3.5 embedding unpacking with TRL vs trainer evaluate-after-train failure." + "reason": "Unpacking Qwen3.5 input embeddings in TRL SFT and trainer.evaluate failing after train are different trainer behaviors." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Both are tokenizer/config related, but one is runtime post-processor updates and the other is save/load of custom token IDs." + "reason": "Special-token post-processor refresh and DiaConfig custom token-ID persistence are related broadly, but not the same concrete bug." }, { "left": "issue:42886", "right": "issue:44843", "accept": true, - "reason": "Same underlying offline tokenizer bug: loading from cache fails under `HF_HUB_OFFLINE=1`, and 44843 pinpoints the unconditional `model_info()` call causing it." + "reason": "Same offline-mode tokenizer/cache failure: both describe HF_HUB_OFFLINE breaking AutoTokenizer because an unconditional hub/model_info lookup is performed." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Exact duplicate: same `Qwen2_5_VLProcessor.apply_chat_template` crash on batched input with `padding=False`." + "reason": "Exact duplicate: same title, same crash location, same batched-input/padding=False failure in Qwen2_5_VLProcessor.apply_chat_template." }, { - "left": "issue:45704", - "right": "issue:45923", + "left": "issue:43421", + "right": "issue:43653", "accept": false, - "reason": "Different bugs: T5/apex memory leak vs missing `supports_gradient_checkpointing` on a trust-remote-code model." + "reason": "Both involve special-token handling, but one is backend post-processor synchronization and the other is BigBirdTokenizer mask-token registration." }, { - "left": "issue:45759", - "right": "issue:45923", + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image Processor void-segmentation-map handling is unrelated to the GenerationMixin import error." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen multimodal bugs: weight_only loading failure vs incorrect vision_position_ids in video input." + }, + { + "left": "issue:44464", + "right": "issue:47050", + "accept": false, + "reason": "Chunked generation with compiled forward and fp16 ONNX attention dtype mismatch are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "SigLIP2 documentation errors are not the same as the HF-vs-JAX implementation discrepancy." + }, + { + "left": "issue:46032", + "right": "issue:46463", "accept": false, - "reason": "Different Qwen/Nemotron model-config issues with no shared failing code path." + "reason": "Mamba2 use_cache correctness is unrelated to a T5 decoder input validation error message." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "Different loading/offloading problems: local checkpoint loading for PEFT adapters vs quantization/offloading behavior." + "reason": "PEFT local-checkpoint loading and offloading with quantization are different failures, even if both involve loading." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Different models and different failures: SigLIP2 wrong outputs vs CamemBERT masked-LM prediction regression." + "reason": "A broken SigLIP2 model output and a CamemBERT masked-LM regression are unrelated model-specific bugs." }, { - "left": "issue:46032", - "right": "issue:46463", + "left": "issue:45704", + "right": "issue:45923", "accept": false, - "reason": "Different bugs: Mamba2Mixer cache misuse vs T5 decoder-input validation error message." + "reason": "T5 memory leak and missing supports_gradient_checkpointing on a trust_remote_code model are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45923", + "accept": false, + "reason": "AutoModelForCausalLM.from_config not unwrapping text_config and a missing gradient-checkpointing flag are different issues." } ] }, "evaluator_result": { - "accept": false, - "feedback": "Mostly grounded, but one claim is a bit overstated: saying 44843 has the \"broadest duplicate value\" is not supported by the packet. The evidence only clearly shows it as the cleaner root-cause report for the offline tokenizer bug, not that it has broader duplicate coverage than the other duplicate pair. The soft-edge verdicts themselves look conservative and reasonable." + "accept": true, + "feedback": "The summary is well grounded in the packet: the cluster is mostly heterogeneous, 44514/44545 is a clear duplicate pair, and 42886/44843 is a plausible soft duplicate based on the offline-cache failure theme. The soft-edge rejections are conservative and avoid overclaiming broader equivalence." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -5650,14 +5615,12 @@ "cluster_id": "cluster-10000-509", "nodes": [ "issue:30064", - "issue:36010", "issue:36331", "issue:39692", "issue:41720", "issue:43295", "issue:43335", "issue:43421", - "issue:43493", "issue:43531", "issue:43540", "issue:43550", @@ -5669,7 +5632,6 @@ "issue:43697", "issue:43720", "issue:43756", - "issue:43782", "issue:43819", "issue:43873", "issue:43927", @@ -5679,7 +5641,9 @@ "issue:44220", "issue:44361", "issue:44488", + "issue:44625", "issue:44661", + "issue:44857", "issue:44898", "issue:44991", "issue:44995", @@ -5696,18 +5660,17 @@ "issue:46832", "issue:46858", "issue:46932", - "issue:47020" + "issue:47020", + "issue:47050" ], "soft_pairs": [ - "issue:39692|issue:43493", - "issue:43421|issue:43653", - "issue:30064|issue:36010", - "issue:43782|issue:45381", "issue:45081|issue:45356", + "issue:44857|issue:47050", "issue:44995|issue:46858", "issue:45356|issue:45907", "issue:43421|issue:44186", "issue:39692|issue:43873", + "issue:39692|issue:47050", "issue:43720|issue:43819", "issue:43295|issue:45834", "issue:43653|issue:43720", @@ -5716,6 +5679,7 @@ "issue:45987|issue:46829", "issue:43540|issue:46294", "issue:43550|issue:44361", + "issue:44625|issue:47050", "issue:45381|issue:46693", "issue:44062|issue:44488", "issue:43596|issue:44220", @@ -5733,232 +5697,226 @@ "issue:44488|issue:47020", "issue:43550|issue:43819" ], - "prepared_review_unit_hash": "3d5e27c89bf88a970488b425d528b1043cb539025fa0b5ab9fb81a066d3c0e5b", + "prepared_review_unit_hash": "05d94adf35a3d606fa491fe4d8f43a221ce8be2162aaeb04a44ace3351517686", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37974, - "estimated_input_tokens": 9494, - "estimated_eval_tokens": 19244 + "soft_pair_count": 32, + "serialized_chars": 37683, + "estimated_input_tokens": 9421, + "estimated_eval_tokens": 19098 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: it mixes unrelated tokenizer, vision/video, cache, config, and performance bugs. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", - "confidence": 0.39, - "canonical_issue_reason": "Issue #43873 is the broadest, most actively discussed open report in the set, so it is the best anchor for this otherwise unrelated cluster.", + "summary": "This set is mostly a grab bag of unrelated bug reports; the soft similarity signals are broad subsystem overlaps (tokenizers, caching, vision/video, quantization) rather than true duplicates. I do not see a single coherent duplicate cluster, and I would reject all listed soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45081 is the best anchor because it is a concrete, code-path-specific tokenizer regression in `_patch_mistral_regex`, and it is the only item with a plausible neighboring report in the same narrow area.", "canonical_pr_reason": null, - "best_issue_reason": "#43873 has the strongest engagement and a regression-style description, making it the most representative issue here.", + "best_issue_reason": "Issue 45081 is the most actionable and narrowly scoped bug report in the set; it describes a specific failure mode with a clear fix surface, unlike the broader or more model-specific reports around it.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:39692", - "right": "issue:43493", - "accept": false, - "reason": "Both are SigLIP2-related, but one is a doc/example failure and the other is an implementation discrepancy; different bugs." - }, - { - "left": "issue:43421", - "right": "issue:43653", - "accept": false, - "reason": "Both involve special tokens, but one is a backend post-processor update issue and the other is a BigBirdTokenizer registration bug." - }, - { - "left": "issue:30064", - "right": "issue:36010", - "accept": false, - "reason": "Completely different subsystems and failures: image segmentation processing vs a GenerationMixin import error." - }, - { - "left": "issue:43782", - "right": "issue:45381", + "left": "issue:45081", + "right": "issue:45356", "accept": false, - "reason": "Different Qwen VL models and different failure modes; no clear shared bug." + "reason": "Both touch Mistral-related tokenizer regex handling, but they describe different failures and model contexts; this is related area, not the same underlying bug." }, { - "left": "issue:45081", - "right": "issue:45356", + "left": "issue:44857", + "right": "issue:47050", "accept": false, - "reason": "Both mention fix_mistral_regex, but one crashes in regex patching while the other is a tokenizer regression with codec handling/warning behavior." + "reason": "Different models and different failure modes: AMP/CUDA loss crash versus fp16 ONNX attention dtype mismatch." }, { "left": "issue:44995", "right": "issue:46858", "accept": false, - "reason": "Both touch cache behavior, but they are different code paths and symptoms: stale indexer cache vs seq_length return-type arithmetic." + "reason": "Both mention cache-like behavior, but one is a stale indexer cache on a second forward pass while the other is a sequence-length tensor issue in static cache prefill arithmetic." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Different areas entirely: tokenizer regression vs a repository API forwarding issue." + "reason": "Tokenizer regression versus a CI helper bug forwarding revision; no shared underlying code-path problem." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer-related, but the failure modes are different and not the same underlying bug." + "reason": "Both are tokenizer issues, but one is about runtime special-token/post-processor updates and the other is a crash in NER/batched padding logic." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "SigLIP2 doc/implementation issues are unrelated to the quantization/offloading regression." + "reason": "Doc-example/model mismatch and quantization failure are unrelated to an offloading behavior regression." }, { - "left": "issue:43720", + "left": "issue:39692", + "right": "issue:47050", + "accept": false, + "reason": "Documentation/model mismatch is unrelated to an ONNX dtype mismatch in attention." + }, + { + "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Different models and different internals; no plausible shared fix." + "reason": "BitNet accelerated loading bug and DAC latent reconstruction bug are different models and different code paths." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Both are model-loading/processing issues, but they affect different models and different failure paths." + "reason": "Processor/tokenizer regression versus long OCR index error in a different model family." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Tokenizer special-token handling and BitNet weight unpacking are unrelated." + "reason": "BigBird special-token decode bug and BitNet packed-weight loading bug are unrelated." }, { "left": "issue:46829", "right": "issue:46832", "accept": false, - "reason": "Both are model output/forward problems, but they concern different architectures and different code paths." + "reason": "Different models and outputs: VideoPrism hidden_states assignment versus RT-DETR feature-level mismatch." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "torch.compile/SDPA failure is unrelated to config save/load token ID loss." + "reason": "torch.compile/SDPA failure in Bamba is unrelated to DiaConfig token ID loss after save/load." }, { "left": "issue:45987", "right": "issue:46829", "accept": false, - "reason": "A cache length contract bug and a video classification output-field bug are not the same issue." + "reason": "Static cache sequence-length return type bug is unrelated to VideoPrism output field assignment." }, { "left": "issue:43540", "right": "issue:46294", "accept": false, - "reason": "Both involve video inputs in Qwen models, but the symptoms and APIs differ enough that they do not look like one bug." + "reason": "Both involve Qwen video input handling, but they are different models and distinct failure conditions." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "Different models and different failures; no common code-path evidence." + "reason": "Bamba compile/SDPA issue and MLuke tokenizer AttributeError are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:47050", + "accept": false, + "reason": "Configuration propagation bug versus ONNX attention dtype mismatch." }, { "left": "issue:45381", "right": "issue:46693", "accept": false, - "reason": "Qwen2.5-VL video position IDs vs FA2 long-context performance regression are unrelated." + "reason": "Qwen video-position-id regression and long-context FA2 performance regression are different problems." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "Both are tokenizer loading issues, but one is an AddedToken constructor error and the other is a specific model load failure." + "reason": "AddedToken special-argument error is unrelated to a tokenizer loading failure for a specific model." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "Deepspeed/BertModel initialization and audio fbank extraction are unrelated." + "reason": "Deepspeed zero3/BertModel index error and fbank feature extraction issue are different code paths." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to AutoConfig missing fields with trust_remote_code." + "reason": "Qwen3-MoE sliding-window behavior and AutoConfig trust_remote_code field loss are unrelated." }, { "left": "issue:43611", "right": "issue:46665", "accept": false, - "reason": "Both involve from_pretrained/model loading, but the reported regressions are different and not the same underlying bug." + "reason": "Base-model-prefix loading breakage and inconsistent from_pretrained behavior in classes are not the same defect." }, { "left": "issue:41720", "right": "issue:44991", "accept": false, - "reason": "CUDA device-mapping assert and tokenizer loading for a specific model are unrelated." + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to tokenizer loading for EMBEDDIA/est-roberta." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio feature extraction and tokenizer mapping-name handling are different subsystems." + "reason": "Audio feature extraction bug and tokenizer mapping-name failure are unrelated." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes and Perceiver image-size handling are unrelated." + "reason": "LayoutLMv2 tokenizer crash and Perceiver interpolation failure are different model-specific bugs." }, { "left": "issue:30064", "right": "issue:36331", "accept": false, - "reason": "Image processing for void segmentation maps is unrelated to a Trainer compute_loss signature regression." + "reason": "Image processor void segmentation-map bug and CustomTrainer signature mismatch are unrelated." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "Special-token registration and DAC latent reconstruction are unrelated." + "reason": "Tokenizer special-token registration bug and DAC latent conversion bug are unrelated." }, { "left": "issue:44008", "right": "issue:46858", "accept": false, - "reason": "Gemma3n variable collision and static-cache seq_length return type are different bugs." + "reason": "Gemma3n variable-name collision in forward is unrelated to static-cache sequence-length arithmetic." }, { "left": "issue:45987", "right": "issue:46932", "accept": false, - "reason": "Cache seq_length contract and Gemma4 quantized embedder dtype casting are unrelated." + "reason": "StaticCache return-type bug and Gemma4 quantized embedder dtype casting bug are unrelated." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "Both are model-specific architecture complaints, but they affect different models and different implementation areas." + "reason": "SwitchTransformers sparse-layer construction bug and Smollm3 RoPE-layer counting bug are different configuration issues." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "CLIP output attentions regression and RT-DETR output differences are unrelated model bugs." + "reason": "CLIP attentions field regression and RT-DETR output differences are distinct model behaviors." }, { "left": "issue:44488", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer load failures, but the root causes differ: one model cannot load and the other passes the wrong type to Unigram." + "reason": "Loading a specific tokenizer model and XLMRoberta tokenizer constructor type mismatch are different issues." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Different model internals and different runtime failures; not a duplicate pair." + "reason": "Bamba compile/SDPA problem and DAC latent mismatch are unrelated." } ] }, @@ -6040,8 +5998,8 @@ "issue:34634|issue:36010", "issue:44448|issue:44671", "issue:44186|issue:44442", - "issue:45183|issue:45479", "issue:45584|issue:45834", + "issue:45183|issue:45479", "issue:43322|issue:45685", "issue:44991|issue:47020", "issue:43644|issue:43749", @@ -6059,190 +6017,190 @@ "issue:46294|issue:46463", "issue:44805|issue:44936" ], - "prepared_review_unit_hash": "38eaa58609d78fb073c39625e6054bb576e3bc86b554712547d6dcc34d2844f9", + "prepared_review_unit_hash": "2b6bf7e879e2a15e1e34a14947da6d5862adb1308748cde14dca78ecb77d139f", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 35863, - "estimated_input_tokens": 8966, - "estimated_eval_tokens": 18188 + "serialized_chars": 35859, + "estimated_input_tokens": 8965, + "estimated_eval_tokens": 18186 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly a collection of unrelated Transformers issues; the soft matches are largely thematic rather than true duplicates. None of the soft pairs look like the same concrete bug/change, so all are rejected. The broadest representative issue is the remote_code regression report.", - "confidence": 0.66, - "canonical_issue_reason": "issue:45020 is the broadest cross-cutting report in the set, covering a class of remote_code regressions rather than a single model-specific bug, so it is the best umbrella canonical issue among these mostly unrelated reports.", + "summary": "This cluster is a loose collection of mostly unrelated Transformers bug reports spanning tokenizers, multimodal models, dtype/loading regressions, and validation/error-message issues. I did not find any pair that clearly shares the same concrete underlying bug, so the soft edges should remain unmerged.", + "confidence": 0.71, + "canonical_issue_reason": "Issue 45020 is the broadest and most cluster-like anchor: it describes a general class of recent-version regressions affecting remote_code models, which is the closest thing here to an umbrella report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the most representative cluster anchor because it is broad, user-facing, and plausibly subsumes adjacent remote_code/config-loading breakages more naturally than the narrower model-specific issues.", + "best_issue_reason": "Issue 45020 is the best single issue to keep as the representative cluster member because it is broad, well-scoped around a recurring regression theme, and most likely to subsume nearby remote_code-related reports without being too narrow.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43643", "right": "issue:44877", "accept": false, - "reason": "Both concern config loading behavior, but one is missing fields with trust_remote_code and the other is strict config rejecting granite_speech; different failure modes and code paths." + "reason": "Both involve config loading, but one is about trust_remote_code returning incomplete configs and the other about strict config blocking granite_speech; different failure modes and code paths." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Both mention _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer AttributeError; different bugs." + "reason": "Both touch _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer attribute crash; not the same bug." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "No meaningful overlap beyond being generic runtime failures; one is AutoImageProcessor-from-URL loading, the other is a template compilation TypeError." + "reason": "Unrelated failures: AutoImageProcessor URL loading versus template-node compilation TypeError." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "Different models and problems: SigLIP2 docs/model-processor mismatch vs Kosmos2.5 OCR long-input indexing." + "reason": "Different multimodal model bugs on different models and paths; only loosely similar on the surface." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Both are multimodal/video-ish, but the actual bugs differ completely: Qwen3OmniMoe video processing vs Gemma3n audio_mel_mask name collision." + "reason": "Completely different model families and error causes (video processing ValueError vs variable-name collision in Gemma3n forward)." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and a GenerationMixin import error are unrelated issues." + "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Both are v5 regressions, but they affect different models and outputs (Pegasus generation vs CamemBERT masked LM)." + "reason": "Both are v5 regression reports, but for different models and different inference behaviors; not a single fixable code-path problem." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "Different tokenizer failures: LayoutLMv2 NER/padding crash vs AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + "reason": "Tokenizer crash on NER/batched padding is unrelated to AutoTokenizer failing to load a specific tokenizer class." }, { - "left": "issue:45183", - "right": "issue:45479", + "left": "issue:45584", + "right": "issue:45834", "accept": false, - "reason": "Input validation error messaging for input_ids/inputs_embeds is separate from the zero-loss single-label classification bug." + "reason": "Empty Whisper transcription after align_special_tokens is a different bug from Kosmos2.5 long OCR input indexing." }, { - "left": "issue:45584", - "right": "issue:45834", + "left": "issue:45183", + "right": "issue:45479", "accept": false, - "reason": "Whisper empty-transcription generation after align_special_tokens is unrelated to Kosmos2.5 long OCR input indexing." + "reason": "One is input validation for input_ids/inputs_embeds, the other is loss computation for num_labels=1 classification; different subsystems." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Segfault loading Llava Next models and an MPS histogram Int-type error in moe are different code paths and failure types." + "reason": "Segfault in Llava Next loading and an MPS histogram dtype error are unrelated." }, { "left": "issue:44991", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer-loading issues, but the titles indicate different tokenizers and different root causes; not the same underlying bug." + "reason": "Both are tokenizer load failures, but for different tokenizers and different root causes; too broad to merge." }, { "left": "issue:43644", "right": "issue:43749", "accept": false, - "reason": "Both involve loading/runtime regressions, but one is junk non-persistent buffers and the other is FSDP CPU RAM-efficient loading; not the same fix." + "reason": "Non-persistent buffer junk filling and FSDP CPU RAM efficient loading are distinct loading/runtime issues." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Device mismatch under CPU offload for Bark is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel bfloat16 inference." + "reason": "Device mismatch under CPU offload is not the same as bfloat16 dtype mismatches in inference." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "Qwen3VL weight_only loading failure and Qwen3.5 flash-attention generation issues are distinct problems." + "reason": "from_pretrained weight_only loading error for Qwen3VL is unrelated to flash-attention generation behavior in Qwen3.5." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Both touch special-token behavior, but one is runtime post-processor updates and the other is Whisper batch_decode skip_special_tokens handling." + "reason": "Special-token post-processor update bug and Whisper batch_decode skip_special_tokens behavior are different APIs and different bugs." }, { "left": "issue:44561", "right": "issue:45020", "accept": false, - "reason": "issue:45020 is a broad umbrella for remote_code regressions, while issue:44561 names one specific root cause; they overlap in theme but not the same concrete bug/change." + "reason": "Related in theme, but 44561 is a specific breakage from removing is_torch_fx_available while 45020 is a broad umbrella report; not the same concrete issue." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer loading for EMBEDDIA/est-roberta and importing pipelines with kernels v0.15 are different dependency/loading failures." + "reason": "Tokenizer loading versus pipelines import failure due to kernels version/revision requirements; different failure paths." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Qwen video-input regression and Kosmos2.5 OCR index error are unrelated multimodal bugs." + "reason": "Video-input regression across Qwen variants and long OCR index error in Kosmos2.5 are unrelated." }, { "left": "issue:43577", "right": "issue:43818", "accept": false, - "reason": "Blip2 dtype staying float32 and Video-LLaVA missing temporal attention/share-weight issues are different model-loading defects." + "reason": "BLIP2 dtype inheritance and Video-LLaVA temporal-attention/weight-sharing are separate model bugs." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "torch.compile+SDPA failure in Bamba is unrelated to dtype mismatches in other models." + "reason": "torch.compile + SDPA failure in Bamba and bfloat16 dtype mismatch reports are not the same code path." }, { "left": "issue:43582", "right": "issue:46512", "accept": false, - "reason": "Apple Silicon TypeError in caching_allocator_warmup and stale config.dtype after from_config(dtype=...) are distinct dtype/utility problems." + "reason": "Apple Silicon TypeError in caching_allocator_warmup and stale model.config.dtype after from_config are distinct dtype-related issues." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "gather_for_metrics truncation of tuple labels and Mamba2Mixer use_cache correctness are different bugs with different code paths." + "reason": "Metric gathering drops labels in the last batch versus Mamba2Mixer use_cache correctness are unrelated." }, { "left": "issue:43825", "right": "issue:43994", "accept": false, - "reason": "An incorrect pipeline error message about translation support is not the same as SigLIP2 nonsensical outputs in AutoModel/pipeline." + "reason": "A misleading pipeline error message and a nonsensical SigLIP2 output bug are different pipeline issues with different root causes." }, { "left": "issue:46294", "right": "issue:46463", "accept": false, - "reason": "Qwen3.5 video feature/token mismatch and T5 decoder input validation messaging are unrelated." + "reason": "Qwen3.5 video chat-template token/feature mismatch and T5 decoder-input validation messaging are unrelated." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Index mask-shape mismatch and trainer.evaluate failing after train are separate training/runtime issues." + "reason": "Mask shape IndexError and trainer.evaluate() failing after train() do not share the same concrete bug." } ] }, @@ -6261,8 +6219,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -6343,14 +6301,14 @@ "issue:43906|issue:46209", "issue:36010|issue:43645" ], - "prepared_review_unit_hash": "c1b39bace777b6dee28f9d2918d679c4ec7a3d0a851551f7ea31c4c43dcc5095", + "prepared_review_unit_hash": "3746c10a6634784cc485ae21840a5883bc350cbe0c6935fd76290630e583883c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, @@ -6365,168 +6323,168 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most items are distinct regressions across unrelated model, tokenizer, loading, and processor code paths. I did not find a convincing duplicate pair. The broadest issue in the set is the remote_code regression report, but it is not a true umbrella for the rest.", - "confidence": 0.34, - "canonical_issue_reason": null, + "summary": "This cluster is not a duplicate set: the soft pairs are mostly superficial matches across unrelated models, loaders, tokenizers, and processors. I would reject every soft edge and keep the issues separate.", + "confidence": 0.95, + "canonical_issue_reason": "No safe canonical issue: the items span many different code paths and models, and none of the soft pairs is a clear same-bug duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the broadest, most central regression report in this set and the best single issue to keep visible, but it does not duplicate the other items.", + "best_issue_reason": "issue:45701 is the broadest, highest-level regression description in the list, so it is the closest thing to a generic representative issue, but it is not a duplicate anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Different model families and failure modes: dtype mismatch in Mistral-3 preprocessing vs an audio_mel_mask attribute collision in Gemma3n forward." + "reason": "Both are multimodal model bugs, but one is an image preprocessor/model dtype mismatch and the other is a Gemma3n variable-collision AttributeError; different code paths and fixes." }, { "left": "issue:43931", "right": "issue:46846", "accept": false, - "reason": "Both are Qwen-related, but one is a load-time weight shape mismatch and the other is a tensor-parallel planning/OOM issue; not the same bug." + "reason": "Both involve Qwen VL models, but one is a load-time weight-shape mismatch and the other is a tensor-parallel planning/OOM issue for a different model/version." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Video-input failures for different Qwen models, but the concrete problems differ: one is a ValueError in Qwen3OmniMoe, the other wrong vision_position_ids in Qwen2.5-VL." + "reason": "Both mention video inputs in Qwen models, but the failures are different: a Qwen3OmniMoe ValueError vs a Qwen2.5-VL rope/indexing bug." }, { "left": "issue:28218", "right": "issue:29159", "accept": false, - "reason": "Both mention tokenizer behavior, but one is about an extra space after added tokens while the other is a broader slow/fast tokenizer inconsistency; not enough to treat as the same defect." + "reason": "Both are tokenizer-related, but one is a specific added-token spacing bug while the other is a broad slow/fast tokenizer inconsistency; not clearly the same defect." }, { "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "Completely different code paths: image processor size mismatch vs bfloat16 dtype mismatches in inference." + "reason": "These are unrelated: OmDet-Turbo image sizing versus dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Both involve model loading, but one is FSDP CPU RAM-efficient loading and the other is init_empty_weights passing an unexpected argument; different regressions." + "reason": "Both concern model loading, but one is FSDP CPU RAM-efficient loading and the other is init_empty_weights / _is_hf_initialized handling; different underlying issues." }, { "left": "issue:43720", "right": "issue:45072", "accept": false, - "reason": "Packed-weight loading failure in BitNet is unrelated to dtype mismatch problems in SwitchTransformers/TimmWrapperModel." + "reason": "Different problems: packed-weight unpacking during accelerate loading versus bfloat16 dtype mismatches." }, { "left": "issue:43645", "right": "issue:46665", "accept": false, - "reason": "Both concern custom model usage, but the failure contexts differ: Jupyter notebook model definition/init vs from_pretrained inside Python classes." + "reason": "Both are about model initialization behavior, but one is a Jupyter/custom-model breakage and the other is from_pretrained inside Python classes; not the same bug." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "Both are about the v5 pipeline removal era, but one is an incorrect error message and the other is stale documentation; not the same underlying bug." + "reason": "One is a runtime error message about unsupported translation tasks; the other is docs still referencing pipeline() after removals. Related theme, but not the same change." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "CamemBERT masked-LM prediction regression is unrelated to a general tokenization-change report." + "reason": "CamemBERT masked-LM output regression and a generic tokenization-version change are not the same concrete failure." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "Blip2 float32 dtype retention and Apple Silicon caching_allocator_warmup TypeError are unrelated issues." + "reason": "These are unrelated: Blip2 dtype propagation versus an Apple Silicon caching_allocator_warmup TypeError." }, { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "Different subsystems and symptoms: BarkProcessor voice_preset vs Trainer compute_loss signature mismatch." + "reason": "No overlap in bug surface: BarkProcessor voice_preset versus CustomTrainer.compute_loss signature mismatch." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "TensorFlow backend packaging and a Mistral tokenizer regex patch crash are unrelated." + "reason": "These touch different subsystems entirely: backend installation extras versus a Mistral tokenizer regex/backend_tokenizer crash." }, { "left": "issue:46175", "right": "issue:46932", "accept": false, - "reason": "Both are Gemma4-related, but one is a massive one-hot position-embedding memory issue and the other is dtype casting for quantized embedders." + "reason": "Both are Gemma4 embedder issues, but one is a huge one-hot position-embedding materialization problem and the other is dtype casting for quantized embedders." }, { "left": "issue:46697", "right": "issue:46728", "accept": false, - "reason": "Both are image-processing bugs, but one is over-padding and the other is an incorrect patch-count calculation; different code paths." + "reason": "Both are image preprocessing bugs, but they affect different models and different calculations (padding vs patch counting)." }, { "left": "issue:45003", "right": "issue:46285", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to CLIP model structure inconsistency." + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to CLIPTextModel structural inconsistency." }, { "left": "issue:43782", "right": "issue:45325", "accept": false, - "reason": "Qwen3VL weight_only loading error and Qwen2.5-VL rope-index scaling bug are different failures." + "reason": "Different Qwen VL bugs: one is from_pretrained weight_only loading, the other is rope/index computation for still-image temporal positions." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and DAC.from_latents/STE mismatch are unrelated." + "reason": "Audio chat-template sampling-rate defaults and DAC.from_latents STE mismatch are different audio-processing problems." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug is unrelated to MXFP4 quantization fallback behavior." + "reason": "Tokenizer special-token registration and MXFP4 quantization fallback are unrelated." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "AppleSilicon allocator warmup TypeError and CUDA AMP crash in LwDetrImageLoss are not the same bug." + "reason": "Apple Silicon allocator warmup and CUDA float16 AMP loss crashes are unrelated failures." }, { "left": "issue:45372", "right": "issue:46326", "accept": false, - "reason": "Both touch mistral_common/generation flows, but one is an import error for ReasoningEffort and the other is a continue_final_message tokenizer bug." + "reason": "Both involve mistral_common integration, but one is an import failure for ReasoningEffort and the other is continue_final_message behavior; distinct code paths." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup failure and PIL backend torchvision dependency regression are unrelated." + "reason": "Float8 storage lookup failure and PIL backend torchvision requirement are completely different bugs." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "KeyError loading MobileLLM is a specific model-loading bug, not the broad remote_code regression reported in 45020." + "reason": "A key error loading MobileLLM-125M is not the same as the broad remote_code compatibility regression." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba-9B-v2 is unrelated to MXFP4 quantization fallback." + "reason": "Bamba torch.compile/SDPA failure and MXFP4 quantization fallback are unrelated." }, { "left": "issue:43906", "right": "issue:46209", "accept": false, - "reason": "A reproduction of an earlier issue and an FP8 inference failure on B200 are not the same underlying bug." + "reason": "An isolated reproduction of an earlier issue is not the same as an FP8 inference failure on B200." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "GenerationMixin import failure and custom-model-in-notebook breakage are different loading/import regressions." + "reason": "GenerationMixin import failure and Jupyter custom-model initialization breakage are different issues, even if both relate to v5 API changes." } ] }, @@ -6545,8 +6503,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -6577,7 +6535,6 @@ "issue:43927", "issue:43931", "issue:44060", - "issue:44291", "issue:44368", "issue:44387", "issue:44479", @@ -6596,7 +6553,9 @@ "issue:45538", "issue:45600", "issue:45759", - "issue:46291" + "issue:45859", + "issue:46291", + "issue:47050" ], "soft_pairs": [ "issue:43526|issue:43697", @@ -6615,6 +6574,7 @@ "issue:42490|issue:43257", "issue:43746|issue:44811", "issue:43854|issue:44977", + "issue:45859|issue:47050", "issue:43643|issue:45759", "issue:38175|issue:44610", "issue:39692|issue:44912", @@ -6625,205 +6585,204 @@ "issue:43329|issue:43450", "issue:43329|issue:44479", "issue:42915|issue:43742", - "issue:34567|issue:40990", - "issue:44291|issue:45042" + "issue:34567|issue:40990" ], - "prepared_review_unit_hash": "93161c8f3e9de1f12545a9063d86d833bf82926ce107ba99bc87c5d92c584709", + "prepared_review_unit_hash": "529a20d51424c14a3b788c9a8524d043f00c5d3c1eff283dc3a94a4103e32bf7", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35908, - "estimated_input_tokens": 8977, - "estimated_eval_tokens": 18210 + "serialized_chars": 36454, + "estimated_input_tokens": 9114, + "estimated_eval_tokens": 18484 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the pairs span unrelated bugs in training, loading, quantization, multimodal/video, audio, and packaging. I don\u2019t see a single underlying duplicate thread that would justify merging any pair.", - "confidence": 0.97, - "canonical_issue_reason": null, - "canonical_pr_reason": null, - "best_issue_reason": null, - "best_pr_reason": null, + "summary": "The set is heterogeneous and does not contain clear duplicate pairs; the soft-similarity matches are mostly same-subsystem, not same-bug, relationships.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43257 is the strongest representative issue here: it describes a concrete, reproducible Qwen3-MoE loading failure with accelerate+deepspeed, has substantial discussion, and is more actionable than the other heterogeneous reports.", + "canonical_pr_reason": "No PRs are present in this cluster.", + "best_issue_reason": "issue:43257 is the best overall issue candidate because it is specific, technically grounded, and appears to capture a real code-path bug rather than a vague symptom or documentation problem.", + "best_pr_reason": "No PRs are present in this cluster.", "soft_edge_verdicts": [ { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "Different models and failure modes: BEiT label reduction bug vs RTDetrV2 output divergence." + "reason": "Different bugs: BeitImageProcessorFast label reduction vs RTDetrV2 output differences under v5." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Both touch Qwen3-MoE, but one is accelerate/deepspeed loading/conversion and the other is a multimodal token-counting video-branch bug." + "reason": "Both involve Qwen/MoE-related code, but one is a loading/conversion bug and the other is a multimodal token-counting video-branch bug." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to the unexpected num_items_in_batch kwarg in compute_loss()." + "reason": "Unrelated failures: missing GenerationMixin import vs CustomTrainer.compute_loss signature mismatch." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "Both are loading-related, but gguf_file breaks from_pretrained flows while EMBEDDIA/est-roberta is a tokenizer-specific v5 regression." + "reason": "Both are tokenizer/loading issues, but they point to different failure modes and different affected code paths." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Different components: Mistral tokenizer regex patch crashes in tokenizer internals, while Gemma4Processor lacks _tokenizer in serve." + "reason": "Different components: Mistral regex patching in tokenizer internals vs Gemma4Processor missing _tokenizer in serve." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "Different lifecycle bugs: deepspeed ZeRO3/non-pretrained Bert index error vs trainer.evaluate() failing after trainer.train()." + "reason": "Distinct training/eval failures with different triggers and symptoms." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Qwen3-MoE sliding_window issue is unrelated to PIL image processors incorrectly requiring torchvision." + "reason": "Qwen3-MoE sliding-window behavior is unrelated to PIL image processors requiring torchvision." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both mention tie-word-embeddings warnings, but they affect different models and setups, not the same code-path defect." + "reason": "Both are warning/tie-weights related, but they concern different models and different root causes." }, { "left": "issue:42175", "right": "issue:46291", "accept": false, - "reason": "Packaging/backend dependency problems, but one is missing TensorFlow in the torch extra and the other is kernels import requiring version/revision." + "reason": "Packaging/backend selection bug vs pipelines import failure due to kernels version/revision requirements." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "DAC.from_latents vs forward-pass mismatch is a model computation bug, while DiaConfig loses custom token IDs across save/load." + "reason": "Different model/config bugs: DAC latent reconstruction mismatch vs DiaConfig token-ID persistence loss." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template compilation error is unrelated to CLIPTokenizer using an absurd model_max_length value." + "reason": "Template compilation error and tokenizer max_length misuse are unrelated." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Video token-counting undefined variables are unrelated to auto_mappings referencing removed Sam3LiteText configs." + "reason": "Video multimodal token counting bug vs stale auto_mappings references to removed configs." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "BEiT reduce_labels bug is unrelated to the MPS attention correctness issue." + "reason": "Separate issues: Beit label reduction vs MPS attention correctness when head dims differ." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Local-path trailing-slash import warning is unrelated to Qwen3 MoE loading with accelerate + deepspeed." + "reason": "Local-path import warning is unrelated to Qwen3-MoE weight conversion during accelerate+deepspeed loading." }, { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "GraniteSpeech PEFT checkpoint loading and Whisper batch_decode skip_special_tokens are different bugs." + "reason": "PEFT/local-checkpoint loading bug vs Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "GLM-4.7-Flash test loading failure is unrelated to Qwen3.5 flash-attention generation problems." + "reason": "GLM-4.7-Flash test loading issue vs Qwen3.5 flash-attention generation regression." + }, + { + "left": "issue:45859", + "right": "issue:47050", + "accept": false, + "reason": "CPU offload incompatibility in Qwen3_5Moe rotary embedding is unrelated to ONNX export dtype mismatch in CanineModel." }, { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Both are config-loading issues, but one drops fields with trust_remote_code and the other fails to unwrap text_config for composite Qwen multimodal configs." + "reason": "AutoConfig trust_remote_code field loss is a different bug from unwrap text_config in composite multimodal configs." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 is unrelated to OmDet-Turbo processor/model input-size mismatch." + "reason": "SigLIP2 zero-probability behavior and OmDet-Turbo processor/model size mismatch are different problems." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 doc example errors and gpt-oss MXFP4 quantization fallback are separate problems." + "reason": "Documentation-example model/processor issues are unrelated to MXFP4 quantization fallback on gpt-oss-20b." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MoE loading with accelerate/deepspeed is unrelated to GPT-oss-20b CUDA OOM." + "reason": "Qwen3-MoE weight conversion failure is not the same as GPT-oss-20b OOM." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio chat-template sampling-rate default bug is unrelated to Whisper batch_decode ignoring skip_special_tokens." + "reason": "Audio chat-template sampling-rate default bug is unrelated to Whisper decode token filtering." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Both are OOM reports, but int4 quantization reserved-memory growth and Sam3Video CUDA OOM are not the same defect." + "reason": "Both mention OOM, but one is an int4 quantization reserved-memory regression and the other is a Sam3Video memory limit issue." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Qwen3 MoE load/convert failure is unrelated to Qwen3-VL weight-shape mismatch on model load." + "reason": "Different Qwen model families and different failures: accelerate/deepspeed loading vs weight-shape mismatch." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Undefined video-token helper variables are unrelated to batched video processor shape errors." + "reason": "Both are video-related, but one is an internal token-counting undefined-symbol bug and the other is a batched output-shape bug in video processors." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "These are both video-related, but one is a token-counting helper bug and the other is a broad v5.3.0 video input regression across several models." + "reason": "Related domain, but not the same bug: undefined helper usage in token counting vs broader v5.3.0 video-input regression across several models." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Qwen3 MoE FineGrainedFP8Config failure is unrelated to MobileLLM key errors when loading facebook/MobileLLM-125M." + "reason": "FineGrainedFP8Config failure in Qwen3Moe is unrelated to a MobileLLM-125M key error." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token counter not updating is unrelated to GPT-oss-20b producing high perplexity on WikiText-2." - }, - { - "left": "issue:44291", - "right": "issue:45042", - "accept": false, - "reason": "init_empty_weights/_is_hf_initialized mismatch is unrelated to PIL image processors requiring torchvision." + "reason": "TrainerState token counter not updating is unrelated to high perplexity on GPT-oss-20b." } ] }, @@ -6868,14 +6827,15 @@ "issue:43824", "issue:43906", "issue:43927", + "issue:44291", "issue:44479", "issue:44568", "issue:44756", "issue:44821", "issue:44857", "issue:44912", - "issue:44936", "issue:44987", + "issue:45042", "issue:45081", "issue:45127", "issue:45137", @@ -6890,16 +6850,16 @@ "issue:46424", "issue:46693", "issue:46701", - "issue:46821", "issue:46832", "issue:46869", "issue:47020" ], "soft_pairs": [ + "issue:44291|issue:45042", "issue:46421|issue:46424", "issue:45459|issue:45803", - "issue:43650|issue:43906", "issue:42915|issue:44821", + "issue:43650|issue:43906", "issue:29159|issue:34567", "issue:43749|issue:44756", "issue:43540|issue:46701", @@ -6921,199 +6881,198 @@ "issue:28282|issue:29127", "issue:42915|issue:45538", "issue:44479|issue:46693", - "issue:45538|issue:46869", - "issue:44936|issue:46821" + "issue:45538|issue:46869" ], - "prepared_review_unit_hash": "cb9e79a650e605f756bf122263826dd3db3dee2e76500741e7a88517a3fad970", + "prepared_review_unit_hash": "d904bdb6356e0b64e8efc00cefb63df8015c8ce4cb6ffae1c74ada682e065872", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35325, - "estimated_input_tokens": 8832, - "estimated_eval_tokens": 17920 + "serialized_chars": 35426, + "estimated_input_tokens": 8857, + "estimated_eval_tokens": 17970 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: most items are unrelated Transformers issues across tokenizers, model loading, training, cache, and packaging. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them.", - "confidence": 0.82, - "canonical_issue_reason": "Issue 45081 is the clearest, most actionable bug report in the set: it names a specific tokenizer-loading code path, gives a concrete exception, and is well-scoped for triage.", + "summary": "This set is highly heterogeneous: it spans unrelated tokenizer, loading, quantization, training, vision, and performance bugs. None of the suggested soft pairs look like true duplicates, so all soft edges are rejected.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 45042 is the clearest, well-scoped, user-facing regression report among an otherwise unrelated set, but there is no strong duplicate cluster here.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45081 is the strongest standalone issue to keep as the representative example because it is concrete, reproducible, and narrowly defined.", + "best_issue_reason": "Issue 45042 is the best anchor only in the sense that it is concrete, reproducible, and clearly describes an actionable bug; it is not actually a duplicate hub for the other items.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Different failures: `init_empty_weights` argument handling vs incorrect torchvision dependency for PIL image processors." + }, { "left": "issue:46421", "right": "issue:46424", "accept": false, - "reason": "Both are cache-related, but one is about lazy cache initialization causing recompilation and the other about max_cache_length only applying to Quantized Cache; different defects." + "reason": "Both mention cache, but one is lazy-init recompilation and the other is `max_cache_length` being honored only for quantized cache." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Both involve bad exception handling, but they affect different functions and different failure modes; not the same bug." + "reason": "Different code paths: tokenizer protobuf error masking vs a bare `except:` in Fuyu tensor conversion." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:42915", + "right": "issue:44821", "accept": false, - "reason": "Unrelated issues: one is a vague data request, the other is a reproduction of a specific upstream bug." + "reason": "Unrelated bugs: FineGrainedFP8Config failure for Qwen3Moe vs loading `AutoImageProcessor` from URL." }, { - "left": "issue:42915", - "right": "issue:44821", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "Different subsystems and symptoms: FP8 config failure vs loading AutoImageProcessor from URL." + "reason": "No overlap in underlying bug; one is a vague placeholder issue, the other is a reproduction for a prior bug." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer inconsistency and TrainerState token counting are unrelated code paths." + "reason": "Tokenizer slow/fast inconsistency is unrelated to `TrainerState.num_input_tokens_seen` not updating." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "Both mention memory/loading, but one is FSDP RAM-efficient loading and the other is mmap/OOM on Strix Halo; not the same root cause." + "reason": "Different symptoms and mechanisms: FSDP CPU RAM efficient loading vs disabling mmap on Strix Halo to avoid OOM." }, { "left": "issue:43540", "right": "issue:46701", "accept": false, - "reason": "Different models and pipelines; the failures are not the same concrete bug." + "reason": "Both involve inputs/pipelines, but one is Qwen3OmniMoe video input validation and the other is a DiffusionGemma pipeline output issue." }, { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "A missing pad_token_id on Llama4Config is unrelated to kernels dependency requirements during pipeline import." + "reason": "Llama4 `pad_token_id` attribute missing is unrelated to kernels v0.15 breaking `transformers.pipelines` import." }, { "left": "issue:28282", "right": "issue:29870", "accept": false, - "reason": "ImportError for missing PyTorch and Bart generate() UnboundLocalError are unrelated." + "reason": "Completely different errors: missing PyTorch ImportError vs BART `decoder_attention_mask` unbound local in `generate()`." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they concern different tokenizers and distinct behaviors." + "reason": "Both are tokenizer regressions, but one is special-token insertion and the other is Kimi-K2.5 codec/fix_mistral_regex behavior." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Loading a specific model in >=5.1.0 and CLIPTokenizer max_length handling are separate bugs." + "reason": "Different tokenizer/model issues: loading `physical-intelligence/fast` vs CLIPTokenizer's `model_max_length` sentinel value." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "One is about tied lm_head weights; the other is about dtype propagation in Blip2 loading." + "reason": "AyaVision weight tying bug is unrelated to BLIP2 dtype staying float32." }, { "left": "issue:43927", "right": "issue:44912", "accept": false, - "reason": "Custom token IDs being lost on save/load is unrelated to MXFP4 quantization fallback." + "reason": "Different bugs: DiaConfig save/load token IDs vs MXFP4 quantization falling back to bf16." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Both are loading failures, but one is a missing PyTorch dependency error and the other is a version-specific model loading regression." + "reason": "Missing PyTorch ImportError is unrelated to Llama 3.2 Vision model loading failure." }, { "left": "issue:43425", "right": "issue:46291", "accept": false, - "reason": "Torch incompatibility and kernels import requirements are different dependency issues with different failure points." + "reason": "Dependency/version problems, but not the same one: Torch 2.10 incompatibility vs kernels v0.15 import failure." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "MacOS bus error in CLIP model loading and TrainerState token accounting are unrelated." + "reason": "MacOS CLIP bus error is unrelated to `num_input_tokens_seen` not updating in TrainerState." }, { "left": "issue:43697", "right": "issue:45357", "accept": false, - "reason": "Different regression surfaces: RTDetrV2 output differences vs incorrect visual encoder keys in save_pretrained." + "reason": "Different underlying bugs: RT-DETR output differences in v5 vs Qwen3.5 save_pretrained visual encoder keys." }, { "left": "issue:44857", "right": "issue:46832", "accept": false, - "reason": "Different RT-DETR bugs: AMP/CUDA loss crash vs forward type error when feature levels exceed backbone outputs." + "reason": "Both are RT-DETR/Lw-DETR related, but one is AMP/CUDA loss crashing and the other is a forward shape/feature-level mismatch." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Both touch weight tying, but one is a missing tie in model construction and the other is model collapse after LoRA merge with extended vocab." + "reason": "Both involve tied weights, but one is missing `lm_head` tying in Mistral3 and the other is LoRA merge with extended vocabulary causing collapse." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Backend packaging/install issue and missing export/import issue are not the same defect." + "reason": "Different issues: missing TensorFlow in a torch install extra vs failing to import `Qwen2_5_VLForConditionalGeneration`." }, { "left": "issue:45081", "right": "issue:47020", "accept": false, - "reason": "Mistral regex patch crash and XLMRobertaTokenizer Unigram constructor error involve different tokenizer implementations and failures." + "reason": "Tokenizer internals differ: Mistral regex/backend tokenizer crash vs XLMRoberta passing a dict to Unigram." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error and an unreleased PEFT version bump are unrelated." + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT minimum-version bump." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "Missing PyTorch dependency ImportError and LayoutLMv3 box-information error are unrelated." + "reason": "Different problems: missing PyTorch dependency vs unclear error message for LayoutLMv3 box content." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "FineGrainedFP8Config failure and CLIPTokenizer max_length issue do not share the same root cause." + "reason": "Qwen3Moe FP8 config failure is unrelated to CLIPTokenizer's `model_max_length` value." }, { "left": "issue:44479", "right": "issue:46693", "accept": false, - "reason": "Video-input regression across models and FA2 long-context performance regression are different problems." + "reason": "One is a video-input regression across Qwen models; the other is a long-context FA2 performance regression." }, { "left": "issue:45538", "right": "issue:46869", "accept": false, - "reason": "CLIPTokenizer max_length default and YOLO multi-GPU device_map failure are unrelated." - }, - { - "left": "issue:44936", - "right": "issue:46821", - "accept": false, - "reason": "Trainer evaluate-after-train failure and TimesFM window_size AttributeError affect different APIs and code paths." + "reason": "CLIPTokenizer configuration issue is unrelated to YOLO multi-GPU device-map loading failure." } ] }, @@ -7132,8 +7091,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -7168,6 +7127,7 @@ "issue:44821", "issue:44849", "issue:44898", + "issue:44936", "issue:44964", "issue:44987", "issue:44991", @@ -7186,6 +7146,7 @@ "issue:46821" ], "soft_pairs": [ + "issue:44936|issue:46821", "issue:43742|issue:44291", "issue:44062|issue:44991", "issue:24643|issue:29159", @@ -7216,218 +7177,218 @@ "issue:43716|issue:46701", "issue:45405|issue:45600" ], - "prepared_review_unit_hash": "03256b278eb183587d9a3f99428992a35f8cb040f10e1be1ed51a3baf6c29f5b", + "prepared_review_unit_hash": "ccbba8c4882103e9064c5e590de88945de8f376985137f5ed097f7b2a84d269c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35830, - "estimated_input_tokens": 8958, - "estimated_eval_tokens": 18172 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36681, + "estimated_input_tokens": 9171, + "estimated_eval_tokens": 18598 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of unrelated transformer regressions and loading errors. The Qwen video-position pair should be treated conservatively: related theme, but not enough evidence here to call it the same underlying bug.", - "confidence": 0.77, - "canonical_issue_reason": "issue:45488 is still the strongest canonical issue candidate: it is open, actively discussed, and describes a concrete tokenizer regression with broader user impact than the other items.", + "summary": "Most pairs are clearly different bugs touching different models/code paths. The only strong duplicate match is the Qwen2.5-VL/Qwen3-VL video-input regression pair, where one issue is a broader family report and the other is a specific manifestation of the same position-id bug.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 44479 is the best canonical issue because it is the broader regression report for the v5.3.0 Qwen video-input bug family and subsumes the more specific Qwen2.5-VL position-id report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45488 is the best issue to anchor the cluster because it is the most active, broadly representative, and user-visible regression in the set.", + "best_issue_reason": "Issue 44479 is the strongest issue anchor in this set: it names the broader affected family, has clear regression framing, and matches the only accepted duplicate edge.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44936", + "right": "issue:46821", + "accept": false, + "reason": "Different failures in different areas: trainer evaluate state vs TimesFM window_size handling." + }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "Both are model-loading failures, but the failure modes and code paths differ (`KeyError` for MobileLLM vs `_is_hf_initialized` in `init_empty_weights`)." + "reason": "Both are load-time TypeErrors, but they involve different mechanisms and different models/code paths." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "Both involve tokenizer loading, but they point to different bugs and models; not the same concrete regression." + "reason": "Both are tokenizer-loading problems, but the symptoms and affected models are unrelated." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "Training/DeepSpeed weight-shape error vs tokenizer consistency issue; unrelated." + "reason": "DeepSpeed weight shape error and tokenizer slow/fast inconsistency are unrelated bugs." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "One is a Qwen3.5 hidden-states bug, the other a broad tokenization change; different symptoms and causes." + "reason": "One is a Qwen3.5 hidden-states bug; the other is a broad tokenization-version complaint." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "Tokenizer/model loading bug vs Apple Silicon allocator TypeError; unrelated." + "reason": "AutoTokenizer/AutoModel gguf handling and Apple Silicon allocator warmup TypeError are different subsystems." }, { "left": "issue:42915", "right": "issue:43749", "accept": false, - "reason": "Different advanced-loading paths (`FineGrainedFP8Config` vs FSDP CPU RAM-efficient loading); not the same bug." + "reason": "Different training/loading failures: FP8 config handling vs FSDP CPU RAM efficient loading." }, { "left": "issue:43577", "right": "issue:46512", "accept": false, - "reason": "Both mention dtype, but one is a Blip2 weight/QFormer dtype mismatch while the other is stale config dtype after `from_config`; distinct issues." + "reason": "Both concern dtype state, but one is model/submodule dtype loading and the other is config.dtype staleness in from_config." }, { "left": "issue:27957", "right": "issue:28282", "accept": false, - "reason": "Flash Attention support vs missing PyTorch import; unrelated." + "reason": "Flash Attention support and missing PyTorch detection are unrelated issues." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Both are Qwen-related runtime failures, but one is auto device mapping/cuda assert and the other is video-input processing in Qwen3OmniMoe." + "reason": "Different Qwen model families and different failure modes: device-mapping CUDA assert vs video input ValueError." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Different Qwen multimodal models and different failure signatures; too far apart to treat as one bug." + "reason": "Both mention video inputs, but they affect different Qwen variants with different errors and likely different fixes." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "A model-loading regression vs a hub API forwarding bug; unrelated." + "reason": "A model load failure and list_repo_files not forwarding revision are unrelated." }, { "left": "issue:42175", "right": "issue:44488", "accept": false, - "reason": "Packaging/backend dependency issue vs model loading failure for a specific checkpoint; unrelated." + "reason": "Package/backend dependency packaging and a model load failure are not the same bug." }, { "left": "issue:44479", "right": "issue:45381", - "accept": false, - "reason": "Both are Qwen video-position regressions, but this packet only supports title-level similarity; there is not enough evidence to conclude the same underlying bug or code path." - }, - { - "left": "issue:43540", - "right": "issue:44479", - "accept": false, - "reason": "Both involve Qwen video paths, but they affect different model families and likely different code paths." + "accept": true, + "reason": "Both report the same Qwen video-input regression in transformers 5.3.0; 45381 is a specific symptom of the broader bug in 44479." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Different vision-model failures with different symptoms; not the same underlying bug." + "reason": "Different vision-model failures: unexpected zero probabilities vs Perceiver resize/interpolation failure." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "Trainer token-count state not updating vs custom `compute_loss` signature change; related to training internals but not the same bug." + "reason": "TrainerState token counting and custom compute_loss signature mismatch are separate trainer issues." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Both are tokenizer regressions, but one is special-token insertion for mDeBERTa and the other is LlamaTokenizer pre-tokenizer override for DeepSeek; distinct causes." + "reason": "Both are tokenizer regressions, but they involve different tokenizers and distinct behavior changes." }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "Model init with `init_empty_weights` vs loading `AutoImageProcessor` from URL; unrelated." + "reason": "Init-empty-weights argument handling and loading AutoImageProcessor from URL are unrelated." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Different multimodal model failures and different operations (test generation vs loading a checkpoint); not duplicate." + "reason": "Different multimodal regressions with different entry points: a Janus test failure vs Phi-4 multimodal load failure." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "MoE weight conversion under accelerate/deepspeed vs RoPE layer count mismatch in Smollm3; unrelated." + "reason": "Qwen3 MoE weight conversion and Smollm3 RoPE-layer dropping are unrelated model-implementation bugs." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer bugs, but one is AddedToken argument handling and the other is Mistral regex patching; different root causes." + "reason": "Different tokenizer bugs: AddedToken special-arg duplication vs Mistral regex patch failure." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "Image-processor dependency regression vs tokenizer regex crash; unrelated." + "reason": "PIL image-processor dependency requirements and Mistral tokenizer regex patching are unrelated." }, { "left": "issue:43582", "right": "issue:46821", "accept": false, - "reason": "Apple Silicon allocator TypeError vs TimesFM window_size attribute error; unrelated." + "reason": "Allocator warmup on Apple Silicon and TimesFM window_size AttributeError are different problems." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "Both are loading-time errors, but the specific failure mechanisms and affected components differ." + "reason": "Different loading failures: MobileLLM key error vs missing Float8 storage object." }, { "left": "issue:43756", "right": "issue:46399", "accept": false, - "reason": "Different Qwen/Smollm model bugs with unrelated mechanisms; not mergeable into one fix." + "reason": "RoPE-layer dropping in Smollm3 and a Qwen3.5 kernel registration crash are unrelated." }, { "left": "issue:43756", "right": "issue:45325", "accept": false, - "reason": "Both touch RoPE/positioning, but one is Smollm3 RoPE-layer count and the other is Qwen2.5-VL temporal position scaling; different bugs." + "reason": "Both concern rotary/position handling, but they affect different models and different bugs." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding-window bug vs image-processor URL loading failure; unrelated." + "reason": "Qwen3-MoE sliding-window behavior and AutoImageProcessor URL loading are unrelated." }, { "left": "issue:43716", "right": "issue:46701", "accept": false, - "reason": "Different multimodal dtype/pipeline issues with no shared concrete code path." + "reason": "Mistral-3 dtype mismatch and DiffusionGemma pipeline failure are different code paths." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "Dependency version pinning vs stale auto-mapping config references; unrelated." + "reason": "PEFT version bump and removed Sam3LiteText auto-mapping references are different maintenance issues." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall summary is mostly grounded, but one soft-edge verdict is not conservative enough: you accepted issue:44479 vs issue:45381 even though the packet\u2019s deterministic label is false. The titles suggest related video-position bugs, but the packet does not provide enough evidence to treat them as the same underlying bug. Please keep that pair rejected or qualify it much more narrowly." + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The one accepted duplicate pair, issue:44479 and issue:45381, is well supported by the titles and the stated v5.3.0 Qwen video-input regression family. The remaining pair rejections are cautious and based on clear differences in model family, subsystem, or failure mode rather than overclaiming equivalence." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -7439,8 +7400,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -7502,8 +7463,8 @@ "issue:24643|issue:29870", "issue:43906|issue:46468", "issue:43540|issue:46693", - "issue:44625|issue:44877", "issue:43452|issue:44488", + "issue:44625|issue:44877", "issue:45663|issue:46846", "issue:43697|issue:46829", "issue:46463|issue:46899", @@ -7521,206 +7482,203 @@ "issue:44493|issue:46050", "issue:43688|issue:43749" ], - "prepared_review_unit_hash": "4b0843a259d52e3d663245f2b488289aaf8878bfdb629fca6d85400608a3565f", + "prepared_review_unit_hash": "37d41222822a9848ef4dbebcee3a9587bc2d720ef470d9db1cb437f8c4bf4b82", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35909, - "estimated_input_tokens": 8978, - "estimated_eval_tokens": 18212 + "serialized_chars": 35916, + "estimated_input_tokens": 8979, + "estimated_eval_tokens": 18214 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue set; the only plausible duplicate pair is the v5 tied-weights/save_pretrained serialization regression. All other soft pairs are different models, different code paths, or different failure modes.", + "summary": "These issues are mostly unrelated one-off bugs across different models, loaders, and training paths. I did not find any soft pair that clearly describes the same underlying bug/change, though a few are loosely similar in theme (e.g. tied-weights or config regressions).", "confidence": 0.78, - "canonical_issue_reason": "Issue 44466 is the clearest anchor for the tied-weights regression: it describes the concrete v5 serialization failure and the device-dependent lm_head.weight behavior directly.", + "canonical_issue_reason": "Issue #45005 is the broadest and most representative report in the set, covering a concrete v5 tied-weights regression with detailed reproduction context; it is the closest thing to a central theme here.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44466 is the strongest representative issue because it is specific about the failing save_pretrained path and generalizes better than the more domain-specific translation-model report.", + "best_issue_reason": "#45005 is the best single issue to anchor this cluster because it is substantial, specific, and captures the most plausible overlap with another tied-weights report without being overly narrow.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44737", "right": "issue:46463", "accept": false, - "reason": "XLNet CPU placement bug vs T5 decoder-input validation message; unrelated symptoms and code paths." + "reason": "Different models and different failures: XLNet CPU device placement vs T5 decoder input validation messaging." }, { "left": "issue:44568", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but one is missing BOS/EOS insertion and the other is leading-space round-trip loss; different bugs." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and symptoms, so they are not the same bug." }, { "left": "issue:45823", "right": "issue:46463", "accept": false, - "reason": "Gemma4 device_map/PLE mismatch is unrelated to T5 decoder-input validation." + "reason": "Gemma4 device-map forward mismatch is unrelated to T5 decoder input validation." }, { "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": "A reproduction of an older issue is not the same as a FA2 long-context performance regression." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Both mention dtype issues, but they affect different models and different inference paths." + "reason": "Both mention dtype issues, but they concern different models and different code paths." }, { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Both are 4-bit quantization problems, but one is a uint8 torch.finfo crash and the other is a bitsandbytes skip-module assertion." + "reason": "Both are 4-bit inference issues, but one is a uint8/torch.finfo crash in Gemma4 audio modules and the other is a BitsAndBytes skip-module exclusion bug." }, { "left": "issue:24643", "right": "issue:29870", "accept": false, - "reason": "DeepSpeed training weight-dimension error is unrelated to Bart generate() unbound variable error." + "reason": "DeepSpeed training weight-shape error and BART generate() unbound variable are unrelated failures." }, { "left": "issue:43906", "right": "issue:46468", "accept": false, - "reason": "Reproduction of an old issue and broken Nemotron example are different problems." + "reason": "An isolated reproduction of one bug is not the same as a broken Nemotron example." }, { "left": "issue:43540", "right": "issue:46693", "accept": false, - "reason": "Video input ValueError in Qwen3OmniMoe is unrelated to FA2 long-context performance regression." + "reason": "Video input processing failure in Qwen3OmniMoe is unrelated to FA2 performance regression." }, { - "left": "issue:44625", - "right": "issue:44877", + "left": "issue:43452", + "right": "issue:44488", "accept": false, - "reason": "Both are config-loading issues, but one is num_labels propagation and the other is strict config loading for granite_speech." + "reason": "Both are loading-related, but one is a gguf_file API break and the other is a specific model load failure; no clear same code-path." }, { - "left": "issue:43452", - "right": "issue:44488", + "left": "issue:44625", + "right": "issue:44877", "accept": false, - "reason": "General gguf_file loading breakage and a specific model load failure are not the same bug." + "reason": "One is num_labels propagation in Qwen3.5, the other is strict config handling for granite_speech." }, { "left": "issue:45663", "right": "issue:46846", "accept": false, - "reason": "FSDP2 shared_kv_states KeyError and missing TP plan entries are different sharding/training failures." + "reason": "Gemma4 FSDP2 shared_kv_states bug and Qwen3.5 TP plan omissions are different distributed-training problems." }, { "left": "issue:43697", "right": "issue:46829", "accept": false, - "reason": "Different models and different output-shape/output-field problems." + "reason": "Different models and outputs: RTDetrV2 v5 regression vs VideoPrism hidden_states field mix-up." }, { "left": "issue:46463", "right": "issue:46899", "accept": false, - "reason": "T5 decoder validation error message is unrelated to Gemma4 multimodal 4-bit dtype casting." + "reason": "T5 decoder input error messaging is unrelated to Gemma4 multimodal dtype casting under 4-bit quantization." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Missing PyTorch import error is unrelated to Trainer _gpu_broadcast_one NoneType failure." + "reason": "Missing PyTorch import and a Trainer broadcast type error are unrelated." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "Both touch config/rope parameters, but one is reload persistence and the other is a JSON list type error in validation utility." + "reason": "Both touch config/rope-related settings, but one is GPTNeoX reload behavior and the other is a utility type error on JSON-loaded lists." }, { "left": "issue:44737", "right": "issue:45588", "accept": false, - "reason": "XLNet CPU arange placement and flash_attention s_aux=None crash are unrelated." + "reason": "XLNet CPU arange placement and flash-attention sink-less model crash are unrelated." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "BEiT label reduction bug is unrelated to single-label-classification zero-loss behavior." + "reason": "BeitImageProcessorFast label reduction and sequence-classification loss selection are unrelated bugs." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Incorrect visual-encoder keys on save_pretrained is unrelated to wrong token IDs from AutoTokenizer." + "reason": "One is Qwen3.5 save_pretrained visual-key serialization; the other is tokenizer token-ID generation across multiple models." }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "RTDetrV2 output regression and Qwen3.5 checkpoint save regression are different models and code paths." + "reason": "RTDetrV2 output drift and Qwen3.5 checkpoint serialization regression are different issues." }, { "left": "issue:45200", "right": "issue:46463", "accept": false, - "reason": "Gemma4 mm_token_type_ids defaulting issue is unrelated to T5 decoder-input validation." + "reason": "Gemma 4 text-only mm_token_type_ids behavior is unrelated to T5 decoder input validation errors." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both involve missing config fields, but one is trust_remote_code AutoConfig loading and the other is a specific Qwen3_5MoeVisionConfig field dropped by strict parsing." + "reason": "Both involve missing config fields, but one is generic trust_remote_code loading and the other is a specific Qwen3_5MoeVisionConfig strict-field issue." }, { "left": "issue:43720", "right": "issue:44898", "accept": false, - "reason": "Packed-weight loading in BitNet and Perceiver interpolation failure are unrelated." + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific bugs." }, { "left": "issue:44466", "right": "issue:45005", - "accept": true, - "reason": "Both report the same v5 tied-weights/save_pretrained regression affecting lm_head serialization; one is the broader symptom, the other a concrete translation-model case." + "accept": false, + "reason": "Both concern tied weights, but one is inconsistent lm_head serialization by device and the other is a broader translation-model regression; not enough evidence they are the same bug." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "Gemma3 ZeRO-3 load IndexError and trainer.evaluate() after train() are different failures." + "reason": "Gemma3 ZeRO-3 loading failure and trainer.evaluate() after train() are unrelated." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Position-id unexpected keys and sinq quantization save/load breakage are unrelated." + "reason": "Unexpected position-id keys and sinq save/load breakage are different serialization problems." }, { "left": "issue:43688", "right": "issue:43749", "accept": false, - "reason": "Aux-loss normalization bug is unrelated to FSDP_CPU_RAM_EFFICIENT_LOADING." + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "Grounded overall. The only accepted soft pair, issue:44466 vs issue:45005, is well-supported by the titles as the same v5 tied-weights / save_pretrained serialization regression, with 44466 as the broader anchor. The other pair verdicts stay conservative and do not overstate similarity from the packet 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 }, @@ -7732,8 +7690,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -7816,202 +7774,202 @@ "issue:44479|issue:44625", "issue:44821|issue:46821" ], - "prepared_review_unit_hash": "d9539a1102b76dd4533ad356ce6c9ee018e3e0298516aa5d0b5a822893f79f51", + "prepared_review_unit_hash": "7fa0c5464532ebcd888ccf091148c103e6e3e01f347c0e43748d8d3117b7bad0", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36298, - "estimated_input_tokens": 9075, - "estimated_eval_tokens": 18406 + "serialized_chars": 36305, + "estimated_input_tokens": 9077, + "estimated_eval_tokens": 18410 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of tokenizer, model-loading, and forward-pass regressions. The soft candidate pairs mostly share broad subsystems but not the same concrete bug, so I rejected all of them.", - "confidence": 0.39, - "canonical_issue_reason": "issue:45701 is the broadest tokenizer-regression report in the set and is the most reasonable umbrella representative for the tokenizer-related complaints here.", + "summary": "This set is a heterogeneous mix of Transformers bugs, with a large tokenizer/regression subtheme but no pair that clearly shares the same concrete root cause. Under a strict duplicate standard, I would reject all soft edges.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45701 is the broadest, most representative report in the set: it describes version-dependent tokenization changes and sits near the center of the tokenizer-related issues, while many others are narrower model-specific regressions.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45701 is the most cluster-representative issue: it is broad, regression-oriented, and closer to the common tokenizer theme than the more model-specific reports.", + "best_issue_reason": "issue:45701 is the best single issue to represent the cluster because it is generic, user-facing, and closest to the recurring tokenizer/regression theme that appears in several other items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Different models and failure modes: zero probabilities in SigLIP vs processor/tokenizer regression in custom model code." + "reason": "Different failure modes: one is zero probabilities for SigLIP2, the other is processor/tokenizer regression with images. Same broad area, not the same bug." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Both affect loading, but one is gguf handling for model/tokenizer load and the other is an offline-mode Mistral regex patch issue." + "reason": "Both involve tokenizer loading, but one is gguf/from_pretrained breakage and the other is offline mode due to an unconditional hub call. Different root causes." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Both are tokenizer load errors, but they target different tokenizer classes and likely different code paths." + "reason": "Both are AttributeErrors in tokenizers, but they affect different tokenizers and loading paths. Not the same underlying issue." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "BigBird special-token registration and MLukeTokenizer task AttributeError are unrelated bugs." + "reason": "BigBirdTokenizer special-token registration/empty decode is unrelated to MLukeTokenizer task-time AttributeError." }, { "left": "issue:43761", "right": "issue:46829", "accept": false, - "reason": "Both mention hidden_states, but one omits them in CLIPVisionModel and the other mislabels outputs in VideoPrism classification." + "reason": "CLIPVisionModel hidden_states regression is a vision-model output bug, while VideoPrism returns the wrong field in an output object. Different code paths." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "Completely different areas: DAC latent reconstruction vs Whisper batch_decode skip_special_tokens handling." + "reason": "DAC latents/STE mismatch is unrelated to Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Both are image-model inference problems, but they affect different models and different failure mechanisms." + "reason": "Both are model inference issues, but SigLIP2 nonsensical outputs and Perceiver interpolation failure have different causes and affected components." }, { "left": "issue:45372", "right": "issue:46170", "accept": false, - "reason": "One is a Gemma/mistral_common import regression; the other is a Windows/ROCm torch.distributed dependency issue." + "reason": "Gemma processor loading import failure and Windows/ROCm distributed import error are distinct dependency problems." }, { "left": "issue:44857", "right": "issue:47000", "accept": false, - "reason": "Both are DETR loss crashes, but the specific losses and root causes differ." + "reason": "LwDetr float16 CUDA crash and RT-DETR cost-matrix infeasibility are separate loss/assignment failures." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Unrelated issues: Qwen3-MoE sliding_window behavior vs MobileLLM loading KeyError." + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM key error are unrelated configuration/loading bugs." }, { "left": "issue:46832", "right": "issue:46932", "accept": false, - "reason": "RT-DETR feature-level mismatch and Gemma4 dtype casting are different bugs." + "reason": "RT-DETR backbone feature mismatch and Gemma4 quantized embedder dtype casting are different model-specific issues." }, { "left": "issue:43525", "right": "issue:46701", "accept": false, - "reason": "Different models and code paths: config pad_token_id missing vs DiffusionGemma pipeline failure." + "reason": "Llama4Config pad_token_id missing and DiffusionGemma pipeline failure do not share a root cause." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but one is processor/tokenizer access in custom model code and the other is Kimi-K2.5 codec/reg-ex handling." + "reason": "Both mention processor/tokenizer regressions, but one is custom processor/tokenizer access and the other is Kimi-K2.5 codec/fix_mistral_regex behavior. Different defects." }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Generic tokenizer inconsistency vs a MacOS bus error in CLIP model loading; not the same bug." + "reason": "Slow vs fast tokenizer inconsistency is a tokenizer API issue; bus error on MacOS with CLIP is a platform crash. No shared bug." }, { "left": "issue:45542", "right": "issue:46178", "accept": false, - "reason": "Both involve backend availability, but tensorboard-without-TF and tensorflow_text backend mapping are distinct issues." + "reason": "Undefined tf backend from tensorboard-only install and BACKENDS_MAPPING tensorflow_text error are separate backend-resolution problems." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Both concern special tokens, but the affected tokenizers and symptoms differ." + "reason": "BigBirdTokenizer special-token registration and mdeberta-v3-base add_special_tokens behavior are both tokenizer issues but not the same tokenizer or fix." }, { "left": "issue:43688", "right": "issue:45704", "accept": false, - "reason": "Aux-loss normalization and apex RMSNorm memory leak are unrelated." + "reason": "Auxiliary loss normalization in MoE models is unrelated to T5 apex FusedRMSNorm memory leak." }, { "left": "issue:43688", "right": "issue:45701", "accept": false, - "reason": "Aux-loss normalization is not a tokenization regression." + "reason": "Auxiliary loss normalization and tokenization-version drift are different subsystems and bugs." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Custom model initialization breakage and FSDP CPU RAM efficient loading are different failures." + "reason": "Transformers 5.0 custom model initialization breakage and FSDP CPU RAM efficient loading are distinct upgrade regressions." }, { "left": "issue:43756", "right": "issue:46402", "accept": false, - "reason": "Smollm3 RoPE-layer loading and Sam3 text encoder weights missing are separate model-specific issues." + "reason": "Smollm3 RoPE-layer mismatch and Sam3 text encoder weights not loading are unrelated model definition/loading issues." }, { "left": "issue:42898", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer behavior changes, but clean_up_tokenization_spaces and CodeLlama leading-space stripping are different regressions." + "reason": "clean_up_tokenization_spaces v5 behavior change and CodeLlama leading-space round-trip regression are both tokenizer-related but different concrete behaviors." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Both relate to generation/state handling, but the compiled-forward chunking issue is not the same as Qwen3.5 recurrent-state reset." + "reason": "Chunked generation with compiled forward and Qwen3.5 recurrent state reset are separate generation/state-management bugs." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Offline model_info patching and EMBEDDIA tokenizer load failure are different tokenizer-loading bugs." + "reason": "Offline-mode failure from an unconditional model_info call and EMBEDDIA tokenizer loading failure are not the same code-path problem." }, { "left": "issue:43761", "right": "issue:47013", "accept": false, - "reason": "Different model families and output-shape bugs; no shared concrete code path." + "reason": "CLIPVision hidden_states regression and xLSTM train-mode unpacking crash are unrelated output/state handling bugs." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "GGUF loading and BLIP2 dtype propagation are unrelated." + "reason": "gguf_file loading breakage and Blip2 dtype propagation are different from-pretrained/model-initialization issues." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "Special-token registration in BigBird and save/load token-id loss in DiaConfig are different tokenization/config bugs." + "reason": "BigBirdTokenizer special-token registration and DiaConfig custom token IDs not persisting are separate tokenizer/config persistence bugs." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Video-input regression in Qwen multimodal models and Qwen3.5 num_labels propagation are not the same change." + "reason": "Video input regression for Qwen3.5-family models and num_labels propagation in Qwen3.5 are different model configuration issues." }, { "left": "issue:44821", "right": "issue:46821", "accept": false, - "reason": "AutoImageProcessor URL loading and TimesFM window_size AttributeError are unrelated." + "reason": "AutoImageProcessor URL loading and TimesFM window_size AttributeError are unrelated loader/config problems." } ] }, @@ -8030,8 +7988,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -8117,220 +8075,220 @@ "issue:44554|issue:46365", "issue:45823|issue:46050" ], - "prepared_review_unit_hash": "8b7ac29e49a8c17a8b2291c68ba384b2ab987a947fe96318ffb8ab66bf278b2d", + "prepared_review_unit_hash": "38cd7d2389c3ef72371ac4f9ea52d9c67e850930c68dc0b65238c25dd0d202c7", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37136, - "estimated_input_tokens": 9284, - "estimated_eval_tokens": 18824 + "serialized_chars": 37142, + "estimated_input_tokens": 9286, + "estimated_eval_tokens": 18828 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab-bag of mostly unrelated Transformers issues: training API breakages, config/load regressions, model-specific generation bugs, and performance/device-mapping problems. None of the soft pairs looks like the same underlying bug, so all soft edges should be rejected.", - "confidence": 0.94, - "canonical_issue_reason": "Issue 43611 is the broadest cross-cutting regression in the set (v5 loading breakage around base_model_prefix) and is the closest thing to a representative anchor, but the overall cluster is still heterogeneous.", + "summary": "Mostly distinct issues with a few thematic overlaps (v5 regressions, chat-template crashes, save/load bugs), but none of the soft pairs look like the same concrete bug/change.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43611 is the clearest broad framework regression in the set: it names a concrete v5 loading breakage with a shared root cause and wider applicability than the model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "43611 is the best representative issue because it has the widest blast radius and is a clear version-wide regression, unlike the other highly model-specific reports.", + "best_issue_reason": "issue:43611 is the best representative issue overall because it is concrete, broadly impactful, and easy to triage against related v5 compatibility regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU Qwen2-VL training and a GenerationMixin import error are different failures in different subsystems." + "reason": "Different failures: multi-GPU Qwen2-VL training vs a GenerationMixin import error." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "A save_pretrained shard-size question and an AutoProcessor kwargs forwarding bug are unrelated." + "reason": "Unrelated topics: shard-size question vs AutoProcessor kwargs forwarding." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "lm_head serialization inconsistency and sinq quantized save/load breakage are different root causes." + "reason": "Both involve save/load, but one is tied-weight serialization and the other is quantized model persistence." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Trainer _gpu_broadcast_one NoneType error is not the same as a Qwen2-VL multi-GPU training issue." + "reason": "Trainer GPU broadcast NoneType error is not the same as Qwen2-VL multi-GPU training support." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Auxiliary-loss normalization and DeepSeekV3/DeepseekV3MoE divergence affect different model math paths." + "reason": "Both are MoE-related, but one is auxiliary-loss normalization and the other is a native-vs-remote implementation mismatch." }, { "left": "issue:44734", "right": "issue:46612", "accept": false, - "reason": "KV-cache continuation indexing in serve is a different bug from beam-search cache reorder skipping." + "reason": "Different generation bugs: KV-cache continuation indexing vs beam-search cache reorder omission." }, { "left": "issue:44977", "right": "issue:46693", "accept": false, - "reason": "Qwen3.5 flash-attention generation correctness and a long-context FA2 performance regression are not the same issue." + "reason": "Flash-attention correctness failure is distinct from a long-context FA2 performance regression." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Qwen3 auto device-mapping cudaErrorAssert and compiled-forward chunked-generation inconsistency are unrelated." + "reason": "Auto device-map cuda assert is unrelated to chunked generation with compiled forward." }, { "left": "issue:44464", "right": "issue:44977", "accept": false, - "reason": "Chunked generation with compiled forward is a different failure mode from Qwen3.5 flash-attention generation." + "reason": "Different symptoms and code paths: compiled-forward inconsistency vs flash-attention generation failure." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Beit legacy-field loading and Bamba torch.compile SDPA failure are separate model bugs." + "reason": "Legacy-field loading bug for BEiT is unrelated to Bamba torch.compile/SDPA failure." }, { "left": "issue:43278", "right": "issue:43688", "accept": false, - "reason": "Embedding dtype drift during train/eval is unrelated to auxiliary-loss normalization." + "reason": "Embedding dtype drift during eval is not the same bug as MoE auxiliary-loss normalization." }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Beit pretrained preset loading and BitNet packed-weight unpacking are different code paths." + "reason": "Pretrained preset loading failure and packed-weight accelerate loading failure are different issues." }, { "left": "issue:43782", "right": "issue:46693", "accept": false, - "reason": "Qwen3VL weight_only load failure and FA2 long-context performance regression do not share a concrete bug." + "reason": "Qwen3VL weight_only load error is different from FA2 long-context performance regression." }, { "left": "issue:45491", "right": "issue:46002", "accept": false, - "reason": "Gemma3 NaN embeddings on mixed-length batches and classification-head NaNs after checkout/init are different symptoms and likely causes." + "reason": "Both mention NaNs, but one is sliding-window attention on GPU and the other is classification-head init/checkout." }, { "left": "issue:43262", "right": "issue:43295", "accept": false, - "reason": "Audio chat-template sample-rate defaulting and processor.tokenizer/image-handling regression are distinct processor bugs." + "reason": "Audio processor sampling-rate default bug is unrelated to processor.tokenizer regression." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Qwen2-VL multi-GPU training and a Trainer compute_loss signature mismatch are different training issues." + "reason": "Multi-GPU training request is not the same as a CustomTrainer compute_loss signature break." }, { "left": "issue:43611", "right": "issue:46846", "accept": false, - "reason": "A broad v5 loading regression is not the same as Qwen3.5 tensor-parallel plan omissions causing OOM/shape errors." + "reason": "Both are v5/model-loading related, but base_model_prefix loading and Qwen3.5 TP-plan linear_attn are separate bugs." }, { "left": "issue:43577", "right": "issue:44554", "accept": false, - "reason": "Blip2 dtype propagation and an MPS attention correctness bug are unrelated model/runtime problems." + "reason": "Blip2 dtype propagation and MPS attention correctness are different failure modes." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "BitNet packed-weight loading and DiaConfig token-id persistence are different serialization/config bugs." + "reason": "Packed-weight accelerate loading and DiaConfig token-ID persistence are unrelated serialization bugs." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Both involve chat templates, but one is batched padding=False handling and the other is tool-call messages with no content; not the same bug." + "reason": "Both hit apply_chat_template, but the triggering conditions and failure paths differ." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch in Bark and DiaConfig token-id loss are unrelated." + "reason": "CPU offload device mismatch is unrelated to DiaConfig save/load token-ID loss." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Beit legacy-field loading and OmDet-Turbo processor/model image-size mismatch are different issues." + "reason": "Legacy-field model loading and OmDet-Turbo processor size mismatch are different issues." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both touch trust_remote_code, but one is missing fields from AutoConfig and the other is removal of is_torch_fx_available; different breakages." + "reason": "Both affect trust_remote_code, but one is missing config fields and the other is removal of a utility import." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Missing pad_token_id in Llama4Config and missing pad_token_idx in StableLmConfig are similar in shape but not the same underlying bug." + "reason": "Similar theme, but different model configs and different missing fields; not one underlying bug." }, { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "Phi-4 multimodal load failure and a not-yet-released peft version bump are unrelated." + "reason": "Model-loading failure and unreleased PEFT-version bump are unrelated." }, { "left": "issue:43720", "right": "issue:44361", "accept": false, - "reason": "BitNet packed-weight loading and MLukeTokenizer AttributeError on tasks are different problems." + "reason": "Different areas: accelerate weight unpacking vs tokenizer AttributeError." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription handling are unrelated." + "reason": "Qwen3.5 num_labels propagation is unrelated to Whisper empty-transcription generation failure." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 flash-attention generation and FP8 inference on B200 are different model/runtime failures." + "reason": "Qwen3.5 flash-attention generation bug is not the same as FP8 inference failure on B200." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 implementation discrepancy and a pipeline error-message regression are unrelated." + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message wording are unrelated." }, { "left": "issue:44554", "right": "issue:46365", "accept": false, - "reason": "An MPS attention correctness issue and a Trainer import conflict with datasets are unrelated." + "reason": "MPS attention correctness and Trainer import failure due to datasets are unrelated." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Gemma4 device-map mismatch and sinq quantized save/load breakage are different bugs." + "reason": "Gemma4 device-map forward mismatch is not the same as quantized model save/load breakage." } ] }, @@ -8349,8 +8307,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -8435,214 +8393,214 @@ "issue:44493|issue:46285", "issue:43504|issue:43653" ], - "prepared_review_unit_hash": "ef3090333c9821018aae2a99161786a24410276e8225db2c12798ba7534d50db", + "prepared_review_unit_hash": "02e888d06bdb2d5e9e494351a1c86b6f1249e92236486e1448218429d3b9e053", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36859, - "estimated_input_tokens": 9215, - "estimated_eval_tokens": 18686 + "serialized_chars": 36856, + "estimated_input_tokens": 9214, + "estimated_eval_tokens": 18684 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly a grab bag of unrelated bug reports; none of the soft pairs appears to be the same underlying issue or change.", - "confidence": 0.32, - "canonical_issue_reason": "Issue 45812 is the clearest broad, user-visible regression with a concrete symptom and wide model impact, making it the strongest anchor issue in this set.", + "summary": "These issues are mostly unrelated regressions across different models/subsystems; despite some tokenizer/loading overlap, none of the soft pairs look like the same underlying bug/change. No PRs are present in the cluster.", + "confidence": 0.97, + "canonical_issue_reason": "issue:45812 is the strongest representative: it describes a concrete, high-impact tokenizer regression with broad scope (all Granite models), a clear version boundary, and a silent behavioral change.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45812 is the most generally useful issue to keep as the representative report: it is specific, high-impact, and describes a clear regression path.", + "best_issue_reason": "issue:45812 is a good global anchor because it is specific, user-facing, and clearly reproducible, making it more representative than narrow model-specific or environment-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer spacing bug vs. MacOS CLIP bus error; different symptoms, models, and failure modes." + "reason": "Tokenizer spacing bug vs MacOS bus error in CLIP loading; different symptoms and code paths." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "Both concern loading models, but one is a gguf/from_pretrained breakage and the other is a specific BERT model compatibility problem." + "reason": "Both are loading failures, but one is gguf/AutoTokenizer/AutoModel behavior and the other is a specific pretrained model compatibility issue." }, { "left": "issue:43994", "right": "issue:46192", "accept": false, - "reason": "SigLIP2 inference quality issue vs. ignored return_tensors handling; not the same code path." + "reason": "SigLIP2 inference quality issue vs return_tensors being ignored in multimodal input handling; not the same bug." }, { "left": "issue:42796", "right": "issue:44977", "accept": false, - "reason": "Tokenizer BPE-marker handling vs. Qwen3.5 flash-attention generation; unrelated bugs." + "reason": "Tokenizer BPE marker handling vs Qwen3.5 flash-attention generation bug; unrelated failures." }, { "left": "issue:39692", "right": "issue:43295", "accept": false, - "reason": "Documentation example errors vs. regression in processor.tokenizer/image passing behavior; different scope and root cause." + "reason": "Documentation example errors vs a regression in processor/tokenizer access; different concrete problems." }, { "left": "issue:43493", "right": "issue:43844", "accept": false, - "reason": "SigLIP2 implementation mismatch vs. ZeRO-3 gradient growth; different model and training-path issues." + "reason": "SigLIP2 JAX parity discrepancy vs a DeepSpeed/ZeRO gradient instability; different layers of the stack." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "Dtype drift in evaluation vs. gradient instability under DeepSpeed; no shared bug." + "reason": "Training/eval dtype mismatch vs gradient explosion under ZeRO-3; not the same underlying defect." }, { "left": "issue:46209", "right": "issue:46710", "accept": false, - "reason": "FP8 inference failure on B200 vs. accuracy regression after upgrade; different problem classes." + "reason": "FP8 inference failure on B200 vs accuracy regression after version upgrade; different hardware/pathology." }, { "left": "issue:44849", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 hidden-states bug vs. FP8/B200 inference failure; unrelated functionality." + "reason": "Qwen hidden-states bug vs FP8 inference issue; no shared code-path evidence." }, { "left": "issue:43646", "right": "issue:44466", "accept": false, - "reason": "Custom model initialization regression vs. serialization of lm_head.weight; different code paths." + "reason": "Custom model initialization breakage vs serialization inconsistency of tied weights; distinct regression types." }, { "left": "issue:44568", "right": "issue:46489", "accept": false, - "reason": "Both tokenizer-related, but one is BOS/EOS insertion for mDeBERTa and the other is DeepSeek-Coder output correctness; not the same bug." + "reason": "mDeBERTa special-token addition bug vs DeepSeek-Coder tokenizer output regression; different tokenizers and causes." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "Training/eval dtype mismatch vs. resume_from_checkpoint key mismatch; unrelated." + "reason": "Embedding dtype mismatch vs resume checkpoint key mismatch; unrelated." }, { "left": "issue:44568", "right": "issue:45812", "accept": false, - "reason": "Different tokenizer regressions affecting different models and behaviors; no evidence they share a single root cause." + "reason": "Both are tokenizer-related, but one concerns special-token addition behavior and the other wrong token IDs for Granite models." }, { "left": "issue:33357", "right": "issue:33666", "accept": false, - "reason": "MacOS bus error on CLIP vs. Qwen2-VL multi-GPU training request; unrelated." + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training support request; unrelated." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "Model loading failure vs. offline-mode regression in Mistral regex patch; different triggers." + "reason": "Model load failure for a specific checkpoint vs offline-mode breakage from unconditional model_info() lookup; different root causes." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "Janus test failure vs. PEFT minimum-version bump; no shared underlying bug." + "reason": "Janus test failure vs a dependency version bump issue; not the same bug." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "Save/load token ID loss in DiaConfig vs. MLukeTokenizer AttributeError; different components." + "reason": "Config save/load losing token IDs vs tokenizer AttributeError on tasks; different failure modes." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Hidden-states bug vs. incorrect visual encoder keys on save_pretrained; unrelated Qwen3.5 issues." + "reason": "Hidden-states computation bug vs incorrect visual encoder key serialization; unrelated Qwen3.5 issues." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "Import error for a missing class vs. float8 storage type lookup failure; unrelated runtime failures." + "reason": "Missing Qwen2.5-VL export/import symbol vs Float8 storage lookup error; distinct API/runtime problems." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "Tensor-parallel OOM vs. PEFT version pinning issue; not the same problem." + "reason": "Tensor-parallel OOM vs unreleased minimum PEFT version bump; unrelated." }, { "left": "issue:46821", "right": "issue:46832", "accept": false, - "reason": "AttributeError in TimesFM vs. RT-DETR forward TypeError; different model-specific failures." + "reason": "TimesFM window_size AttributeError vs RT-DETR feature-level mismatch; different model families and bugs." }, { "left": "issue:45245", "right": "issue:46710", "accept": false, - "reason": "Category cardinality runtime limit vs. accuracy regression after upgrade; no common root cause." + "reason": "Category-limit runtime error vs post-upgrade accuracy regression; no shared defect." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "Missing config field after 5.0 update vs. unexpected init argument with init_empty_weights; both version-related but distinct bugs." + "reason": "Missing config field after 5.0 update vs init_empty_weights constructor argument mismatch; separate upgrade regressions." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Missing pad_token_id in Qwen3-VL vs. incorrect unwrap of text_config in composite multimodal configs; related area, but different defects." + "reason": "Missing pad_token_id on Qwen3-VL text config vs text_config unwrapping for composite causal LM configs; related family, but not the same concrete bug." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Legacy field prevents BEiT loading vs. CPU offload device mismatch in bark-small; unrelated." + "reason": "Legacy field loading issue in BEiT vs device mismatch with CPU offload in bark-small; unrelated." }, { "left": "issue:44589", - "right": "issue:45520", + "right": "issue:46285", "accept": false, - "reason": "Float8 storage lookup failure vs. flash_attn key error on Python 3.13; different compatibility problems." + "reason": "Float8 storage type lookup vs CLIP text-model structure inconsistency; different subsystems." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Both are config-field regressions, but they affect different models and fields; not the same underlying bug." + "reason": "Both mention missing config fields, but one is pad_token_id on multimodal text config and the other is pad_token_idx in StableLmConfig; not the same bug." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Qwen3 MOE conversion/loading path vs. Qwen3VL MOE model loading/import path; related family, but different breakages." + "reason": "Different Qwen3 MOE loading regressions with different triggers and errors." }, { "left": "issue:44493", "right": "issue:46285", "accept": false, - "reason": "Unexpected position-id key warnings vs. CLIP text-model structure inconsistency; different model families and failures." + "reason": "Unexpected position-id key warnings across many models vs CLIP text-model structure change; not one underlying issue." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "Pretrained loading with a legacy field vs. BigBirdTokenizer special-token registration/empty decode bug; unrelated." + "reason": "BEiT legacy-field load failure vs BigBirdTokenizer special-token registration bug; unrelated." } ] }, @@ -8661,8 +8619,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -8727,8 +8685,8 @@ "issue:45588|issue:46050", "issue:45200|issue:45229", "issue:45229|issue:45823", - "issue:29159|issue:32090", "issue:29942|issue:30333", + "issue:29159|issue:32090", "issue:42490|issue:43696", "issue:45084|issue:45520", "issue:28218|issue:34567", @@ -8742,190 +8700,184 @@ "issue:44977|issue:46710", "issue:37428|issue:43957" ], - "prepared_review_unit_hash": "beafc55f5dded26c8c51e716734783f271c8afc01892e9be5a9329e513e1ebab", + "prepared_review_unit_hash": "fb940c46b71563f544980be370db330ca8e2c1d3fdcfc25a49bf11e9e3a36f00", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 34904, - "estimated_input_tokens": 8726, - "estimated_eval_tokens": 17708 + "serialized_chars": 34913, + "estimated_input_tokens": 8729, + "estimated_eval_tokens": 17714 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a heterogeneous grab-bag of unrelated transformer/model bugs (tokenization, loading, quantization, device mapping, attention, serving). None of the proposed soft pairs look like the same underlying defect or fix.", - "confidence": 0.97, - "canonical_issue_reason": "issue:43873 is the broadest and most actively discussed issue here, so it is the best representative anchor if one is needed; however, this cluster does not contain a true duplicate issue group.", + "summary": "The soft pairs are only broad thematic similarities; none look like the same underlying bug or change, so all candidate merges are rejected. This cluster is heterogeneous and should not be deduplicated into a single issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43873 is the best available anchor only because it is open and describes a broad runtime/offloading regression, but the cluster is too heterogeneous for true duplicate canonicalization.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the most representative single issue in the set because it is broad, active, and has the most inbound references; it is still only a loose thematic anchor, not a duplicate canonical.", + "best_issue_reason": "issue:43873 is the most reasonable triage anchor for the cluster, but it should not be treated as a duplicate magnet because the other issues span unrelated models and failure modes.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46399", "right": "issue:46846", "accept": false, - "reason": "Both are Qwen3.5-related, but one is a kernel registration error and the other is a tensor-parallel plan/OOM issue; different code paths and fixes." + "reason": "Different bugs: one is a missing decorator/registration crash, the other is missing TP-plan entries causing OOM/shape errors." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "Different models and failure modes: CPU offload device mismatch vs a stale GraniteSpeech CI test." + "reason": "Unrelated: CPU offload device mismatch for Bark vs a GraniteSpeech CI test failure." }, { "left": "issue:45072", "right": "issue:46832", "accept": false, - "reason": "One is a bfloat16 dtype mismatch in inference; the other is an RT-DETR backbone/feature-level shape issue." + "reason": "Different code paths and models; dtype mismatch in inference is not the same as RTDetr forward shape handling." }, { "left": "issue:44464", "right": "issue:46424", "accept": false, - "reason": "Chunked compiled generation inconsistency and cache max length handling are related to generation, but they are not the same bug or fix." + "reason": "Chunked generation inconsistency with compiled forward is distinct from cache max-length handling for quantized cache." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Quantization/offloading runtime behavior vs broken save_pretrained/from_pretrained serialization; distinct problems." + "reason": "One is quantization/offloading behavior, the other is broken save/load for sinq-quantized models." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "ZeRO-3 gradient growth during training is unrelated to a model load failure for Phi-4 multimodal." + "reason": "ZeRO-3 gradient growth during training is unrelated to loading a specific multimodal model on latest transformers." }, { "left": "issue:43994", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 output correctness and unexpected position-id keys are different symptoms with different likely causes." + "reason": "SigLIP2 output corruption is a different symptom than the position-id key warning affecting many models." }, { "left": "issue:43937", "right": "issue:46701", "accept": false, - "reason": "GLM-5 GenerationConfig validation and DiffusionGemma pipeline failure are unrelated model-specific issues." + "reason": "Invalid GenerationConfig for GLM-5 is unrelated to DiffusionGemma pipeline failure." }, { "left": "issue:44869", "right": "issue:46752", "accept": false, - "reason": "Whisper timestamp decode crash and empty-chat-template crash affect different tokenizer utilities and code paths." + "reason": "Whisper timestamp decoding crash and empty chat-template input crash are different functions and failure modes." }, { "left": "issue:45588", "right": "issue:46050", "accept": false, - "reason": "Flash-attention s_aux handling is unrelated to quantized model save/load corruption." + "reason": "Flash-attention sinkless-model AttributeError is unrelated to quantized model save/load corruption." }, { "left": "issue:45200", "right": "issue:45229", "accept": false, - "reason": "Gemma4 text-only token-type defaults and Gemma4 multi-GPU OOM are different bugs, not one shared fix." + "reason": "Text-only mm_token_type_ids defaulting is a different issue from multi-GPU inference OOM." }, { "left": "issue:45229", "right": "issue:45823", "accept": false, - "reason": "Both involve Gemma4, but one is an OOM and the other is a device mismatch in PLE forward execution." + "reason": "Both involve Gemma4, but one is inference OOM and the other is a device mismatch in forward; not the same bug." }, { - "left": "issue:29159", - "right": "issue:32090", + "left": "issue:29942", + "right": "issue:30333", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to a Trainer broadcast TypeError." + "reason": "Flash Attention 2 test failures do not match MLFlow reporting after training-job failure." }, { - "left": "issue:29942", - "right": "issue:30333", + "left": "issue:29159", + "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and MLflow failure-reporting behavior are unrelated." + "reason": "Tokenizer slow/fast inconsistency is unrelated to Trainer NoneType broadcast errors." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "A trailing-slash import warning and CUDA OOM on GPT-oss-20b are not the same underlying issue." + "reason": "Local-path import warning has nothing to do with GPT-oss CUDA OOM." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Template-node compilation failure and Python 3.13 flash_attn import key error are distinct implementation bugs." + "reason": "Template-node compilation error is unrelated to the flash_attn import KeyError on Python 3.13." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Added-token spacing behavior and TrainerState token counting are unrelated tokenizer/training concerns." + "reason": "Tokenizer added-space behavior and TrainerState token counter updates are different subsystems and bugs." }, { "left": "issue:43065", "right": "issue:46369", "accept": false, - "reason": "SAM3 pixel decoder dummy layer and SAM3-litetext memory overflow are only loosely model-related, not the same defect." + "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as memory overflow in sam3-litetext." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript GeneratorExp support and SigLIP2 zero-probability outputs are unrelated issues." + "reason": "TorchScript GeneratorExp export failure is unrelated to SigLIP2 zero-probability outputs." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "MPS attention correctness and a PyTorch version check bug in flex_attention are different concerns." + "reason": "MPS attention correctness on mismatched head dims is distinct from an incorrect PyTorch version gate." }, { "left": "issue:44977", "right": "issue:46701", "accept": false, - "reason": "Qwen3.5 flash-attention generation failure and DiffusionGemma pipeline failure are different model/code-path problems." + "reason": "Qwen3.5 flash-attention generation bug is unrelated to DiffusionGemma pipeline failure." }, { "left": "issue:41720", "right": "issue:46710", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and DeepSeek accuracy regression are unrelated." + "reason": "Qwen3 auto device-map CUDA assert and DeepSeek accuracy regression are different problems and likely different code paths." }, { "left": "issue:43493", "right": "issue:43701", "accept": false, - "reason": "SigLIP2 HF/JAX discrepancy and resume checkpoint key mismatch are different kinds of bugs." + "reason": "SigLIP2 JAX/HF discrepancy is unrelated to resume_from_checkpoint key mismatch." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing are both serve-related, but not the same bug." - }, - { - "left": "issue:44977", - "right": "issue:46710", - "accept": false, - "reason": "Flash-attention generation issue in Qwen3.5 and DeepSeek regression after upgrade are unrelated." + "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing crash are both serve bugs but not the same underlying defect." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Flash-support import error and meta-device loading breakage are separate compatibility/loading problems." + "reason": "Flash attention utility import error is unrelated to meta-device model loading breakage." } ] }, @@ -8944,8 +8896,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -9030,14 +8982,14 @@ "issue:43824|issue:43975", "issue:43278|issue:43856" ], - "prepared_review_unit_hash": "65faf0d68474511ce5730f61b37f35903cf9979c96762d22c737bea74731b135", + "prepared_review_unit_hash": "41286fed517910d66e98a92f69332d114a6cb8f56938a2c97df940e47f92f21c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, @@ -9052,192 +9004,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab bag of unrelated issues across tokenization, generation, multimodal processors, import paths, quantization, and dtype/device behavior. None of the soft pairs look like the same underlying bug, and there are no PRs to consolidate.", - "confidence": 0.98, - "canonical_issue_reason": "No clear canonical issue; the cluster spans many unrelated failures, so there is no single underlying bug to anchor the cluster.", + "summary": "No soft-edge pair looks like the same underlying bug/change; most are only loosely related by broad subsystems like tokenization, multimodal processors, or generation.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43295 is the strongest representative of the set: a concrete, user-facing regression in core processor/tokenizer handling for multimodal inputs, with clearer reproduction and broader impact than the other issues.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong global representative; the soft edges are all false positives across different subsystems and failure modes.", + "best_issue_reason": "issue:43295 is the best global issue candidate because it is a real regression with explicit breakage in a central API path, rather than a narrow model-specific or doc-only problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing `_tokenizer` and rope-validation TypeError are unrelated failures." + "reason": "Gemma4Processor missing _tokenizer and a rope-validation TypeError are unrelated bugs." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Processor/tokenizer regression vs batched video-shape bug; different code paths." + "reason": "Both involve multimodal processor code, but one is a processor.tokenizer regression and the other is a batched video-shape bug." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "bfloat16 inference dtype mismatch is unrelated to Wav2Vec2 save/tokenization failure." + "reason": "Different models and different failures: inference dtype mismatch vs save_pretrained/tokenization failure." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention sink-less crash and `AutoModelForCausalLM` text_config unwrapping are different bugs." + "reason": "Flash-attention sink-less-model crash and composite-config text_config unwrapping are different code paths." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError is unrelated to Gemma4 mm_token_type_ids defaults." + "reason": "Different models and different symptoms; no shared underlying bug." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor/tokenizer regression and quantization offloading behavior are not the same defect." + "reason": "Processor/tokenizer regression is unrelated to quantization offloading behavior." }, { "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "LayoutLMv2 NER/batching crash vs `return_tensors` being ignored; different failure modes." + "reason": "Both are tokenizer-adjacent, but one crashes on LayoutLMv2 NER/padding and the other ignores return_tensors with text_kwargs." }, { "left": "issue:43572", "right": "issue:46821", "accept": false, - "reason": "StableLmConfig `pad_token_idx` regression is unrelated to TimesFM `window_size` AttributeError." + "reason": "StableLmConfig pad_token_idx and TimesFM window_size are unrelated configuration errors." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device mismatch and `return_tensors`/`text_kwargs` handling are distinct issues." + "reason": "Gemma4 device-map mismatch and return_tensors being ignored are different issues." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "Custom processor/tokenizer regression vs OmDet input-size mismatch; not the same bug." + "reason": "Both are multimodal processor/model mismatches, but they affect different model pipelines and different failure modes." }, { "left": "issue:43550", "right": "issue:43844", "accept": false, - "reason": "Bamba SDPA `torch.compile` failure is unrelated to ZeRO-3 gradient inflation." + "reason": "torch.compile+SDPA failure is unrelated to ZeRO-3 gradient inflation." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Defaulting `mm_token_type_ids` to zeros is different from Gemma4 PLE device placement." + "reason": "Same model family, but one is a missing default tensor field and the other is a device placement bug." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "Python 3.13 `flash_attn` import KeyError and TensorFlow Text backend mapping error are unrelated." + "reason": "Different import/backend registry failures with no shared bug path." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Batched video-shape bug and Qwen2-57B TP OOM are not the same problem." + "reason": "Video batching shape bug and tensor-parallel OOM are unrelated." }, { "left": "issue:29870", "right": "issue:29942", "accept": false, - "reason": "Bart `generate()` decoder mask error and Flash Attention 2 test failures are different issues." + "reason": "Bart generate UnboundLocalError and Flash Attention 2 test failures are not the same bug." }, { "left": "issue:43572", "right": "issue:44987", "accept": false, - "reason": "StableLmConfig missing pad token index and a model load failure are unrelated." + "reason": "Config pad_token_idx regression and loading failure for a specific model are different problems." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Missing flash-attention helper import vs Qwen2.5-VL class import error; different symbols and causes." + "reason": "Both are ImportErrors, but they miss different symbols from different modules." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Generation kwargs sync bug is unrelated to Qwen3.5 `output_hidden_states` behavior." + "reason": "Generation kwargs sync issue and output_hidden_states behavior on Qwen3.5 are distinct." }, { "left": "issue:39692", "right": "issue:46192", "accept": false, - "reason": "SigLIP2 docs/example errors are unrelated to `return_tensors` handling." + "reason": "Documentation example errors are not the same as tokenizer argument handling." }, { "left": "issue:42490", "right": "issue:43493", "accept": false, - "reason": "Trailing-slash local model warning and SigLIP2 implementation discrepancy are not the same bug." + "reason": "Local-path import warning and SigLIP2 implementation discrepancy are unrelated." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "All-zero multimodal assistant masks and tool-call/no-content crashes are distinct chat-template bugs." + "reason": "Both touch apply_chat_template, but one is assistant-mask generation and the other is a crash on tool-call messages." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "MPS attention value-head dim bug and Kimi tokenizer codec regression are unrelated." + "reason": "MPS attention correctness and Kimi tokenizer regression are unrelated." }, { "left": "issue:43295", "right": "issue:43493", "accept": false, - "reason": "Processor/tokenizer regression and SigLIP2 JAX parity discrepancy are different code paths." + "reason": "Processor/tokenizer regression and SigLIP2 HF-vs-JAX discrepancy are different issues." }, { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Qwen2.5-VL video position IDs and FP8 inference on B200 are unrelated issues." + "reason": "Qwen2.5-VL video position IDs and FP8 inference on B200 are different bugs." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen2.5-VL import error are distinct." + "reason": "Qwen3Moe FP8 failure and a Qwen2.5-VL import error are unrelated." }, { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Serialization of tied weights and LoRA merge collapse are not the same defect." + "reason": "Both involve tied weights, but serialization inconsistency and LoRA-merging collapse are not the same underlying bug." }, { "left": "issue:44737", "right": "issue:46894", "accept": false, - "reason": "XLNet CPU tensor creation bug and CohereASR double-shift loss bug are unrelated." + "reason": "XLNet CPU tensor creation and CohereASR loss shifting are unrelated." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Extra tokenizer spacing after added tokens and post-init reinitialization of output embeddings are different behaviors." + "reason": "Tokenizer spacing after added tokens and post_init reinitializing output embeddings are different bugs." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Qwen2.5-VL import failure and DeepSeek Coder detokenization regression do not share a concrete cause." + "reason": "ImportError for a missing class and DeepSeek detokenization regression are unrelated." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype drift in evaluate and Qwen3 MoE memory usage are unrelated." + "reason": "Training/eval dtype drift and Qwen3 MoE memory usage are different problems." } ] }, @@ -9256,8 +9208,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -9342,214 +9294,214 @@ "issue:37428|issue:43425", "issue:28282|issue:33357" ], - "prepared_review_unit_hash": "d6701344f6ad445a83cb9624f26376f91fdfb7c6e8549c9d0e438a1a0e5494c4", + "prepared_review_unit_hash": "534f0b749d3c0e7e0129bcc58ec4e4248d6501869d6d9af2456cdc46f9e5063b", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 37087, - "estimated_input_tokens": 9272, - "estimated_eval_tokens": 18800 + "serialized_chars": 37093, + "estimated_input_tokens": 9274, + "estimated_eval_tokens": 18804 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: the issues span unrelated models, processors, cache paths, docs, and environment failures. I don\u2019t see any soft edge that looks like the same underlying bug/change, so no pair should be merged as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "#43950 is the broadest core regression in the set (a generic `from_pretrained()` load-path corruption issue) and is the best single representative if one must be chosen.", + "summary": "This set does not form a coherent duplicate cluster: the issues span unrelated bugs across loading, processors, generation, caching, and model-specific regressions. No soft edge looks like the same underlying defect, so the cluster should be split rather than merged.", + "confidence": 0.93, + "canonical_issue_reason": "Issue #43950 is the broadest core regression in the set: a from_pretrained loading corruption bug that can affect many models and code paths, making it the best representative if one issue must anchor the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "#43950 is the most globally useful issue to anchor the cluster because it describes a cross-cutting, high-impact regression rather than a model-specific edge case.", + "best_issue_reason": "#43950 is the most central/high-impact issue among mostly model-specific reports, and its scope is wide enough to represent the cluster\u2019s loader/regression theme.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43452", "right": "issue:46512", "accept": false, - "reason": "Different failure modes: GGUF loading vs stale `config.dtype` after `from_config(dtype=...)`." + "reason": "Different failures: gguf loading path vs stale config dtype after from_config; no shared underlying bug." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "Tokenizer special-token decoding bug vs CI stale device-override test; unrelated code paths." + "reason": "Tokenizer special-token registration and a CI device-override failure are unrelated." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat template sampling-rate default vs batched video processor output shape; not the same bug." + "reason": "Audio chat-template sampling-rate default vs batched video shape bug; different subsystems and causes." }, { "left": "issue:43950", "right": "issue:46285", "accept": false, - "reason": "Both touch model loading/structure, but one is a generic buffer-corruption regression and the other is CLIP text-model API inconsistency." + "reason": "Both touch loading/model structure, but one is non-persistent buffer corruption and the other is CLIP class API inconsistency." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Same symptom class (`pad_token_id` missing), but different model configs and separate fixes." + "reason": "Similar symptom name, but different model configs and likely separate fixes." }, { "left": "issue:45200", "right": "issue:46899", "accept": false, - "reason": "Gemma4 text-only multimodal token IDs vs Gemma4 4-bit quantized embedder dtype casting; distinct bugs." + "reason": "Gemma4 text-only token-type-id defaulting and Gemma4 quantized embedder dtype casting are distinct bugs." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "Tokenizer loading failure vs save/tokenization failure in a different model family." + "reason": "Two separate speech-tokenizer failures for different models and code paths." }, { "left": "issue:45200", "right": "issue:46932", "accept": false, - "reason": "Input-token defaults vs quantized vision/audio embedder casting; same model family but different underlying defect." + "reason": "The later Gemma4 issue explicitly references another fix; it is not the same as the token_type_ids bug." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Perceiver resolution handling vs RT-DETR memory-release behavior; unrelated." + "reason": "Image-size/interpolation failure and RT-DETR memory cleanup are unrelated." }, { "left": "issue:43526", "right": "issue:46192", "accept": false, - "reason": "Different processor bugs: label reduction vs `return_tensors` being ignored." + "reason": "Processor label reduction bug vs return_tensors being ignored; different API layers and symptoms." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "Janus image-generation test failure vs stale `num_return_sequences` warning in paged generate; unrelated." + "reason": "Janus image generation test failure is not the same as a stale warning in paged generate()." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Different multimodal issues: Gemma4 token-type defaults vs Qwen2.5-VL temporal position scaling." + "reason": "Both are multimodal issues, but one is token-type IDs and the other is rope position scaling; different defects." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent mismatch vs CI stale device-override test; not the same code-path problem." + "reason": "DAC latent reconstruction mismatch and a GraniteSpeech CI test issue are unrelated." }, { "left": "issue:41720", "right": "issue:46209", "accept": false, - "reason": "Both are hardware/runtime failures, but one is auto device mapping on A800 and the other is FP8 inference on B200." + "reason": "Different failure modes: A800 cudaErrorAssert during auto device mapping vs FP8 inference failure on B200." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "SwitchTransformers config construction bug vs Bamba torch.compile/SDPA failure; unrelated." + "reason": "SwitchTransformers sparse-layer config bug and Bamba torch.compile+SDPA failure do not share a code-path." }, { "left": "issue:44561", "right": "issue:46234", "accept": false, - "reason": "Both involve missing utility helpers, but they are different symbols with different breakages and fixes." + "reason": "Missing import_utils helper after v5 removal vs a different missing availability helper; not the same bug." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs issues about pipelines, but they concern different APIs and regressions." + "reason": "Both are docs-related, but one is pipeline removal references and the other is text-classification return_all_scores behavior." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "Different vision-model bugs: processor input size mismatch vs AMP crash in loss computation." + "reason": "OmDet processor input-size mismatch and LwDetr AMP crash are separate model-specific bugs." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "DAC STE mismatch vs qwen3_omni_moe feature-length helper behavior; unrelated." + "reason": "DAC latents mismatch and qwen3_omni_moe length helper behavior are unrelated." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "Both mention labels, but one drops elements in metric gathering while the other reduces labels incorrectly in an image processor." + "reason": "Both involve label handling, but one is gather_for_metrics truncation and the other is BeitImageProcessorFast reduce_labels behavior." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "CPU offload device mismatch vs feature-length helper arithmetic; unrelated." + "reason": "CPU offload device mismatch and an audio length helper bug are unrelated." }, { "left": "issue:44857", "right": "issue:45593", "accept": false, - "reason": "Different detection-model problems: float16 AMP crash vs missing auxiliary losses." + "reason": "Both are vision-related, but one is AMP instability and the other is missing auxiliary losses when denoising is off." }, { "left": "issue:46512", "right": "issue:46821", "accept": false, - "reason": "Stale `config.dtype` vs `window_size` attribute error; different subsystems." + "reason": "Stale config dtype after from_config is not the same as TimesFM window_size attribute errors." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "Audio feature-extractor arg validation vs Whisper empty-transcription generation failure; not the same bug." + "reason": "Audio feature-extractor center-arg crash and Whisper empty-transcription failure after token alignment are different bugs." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "Siglip2 zero-probabilities issue vs RT-DETR memory leak; unrelated." + "reason": "Unexpected zero probabilities in SigLIP2 and RT-DETR memory retention are unrelated." }, { "left": "issue:45759", "right": "issue:46285", "accept": false, - "reason": "Composite Qwen config unwrapping bug vs CLIP text-model structural inconsistency; different code paths." + "reason": "Composite Qwen CausalLM config unwrapping vs CLIP text-model structure mismatch are different model-API issues." }, { "left": "issue:46424", "right": "issue:46858", "accept": false, - "reason": "Cache configuration coverage issue vs static-cache seq-length tensor arithmetic bug; distinct defects." + "reason": "Cache length config limitation and static-cache slice arithmetic on non-CUDA backends are separate cache bugs." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "SwitchTransformers config bug vs BigBird tokenizer special-token registration bug; unrelated." + "reason": "Sparse-layer config creation and tokenizer special-token registration are unrelated." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "Specific import error for `_flash_supports_window_size` vs a broad torch-version incompatibility report; not the same change or bug." + "reason": "Missing internal flash-attention symbol and broad Torch 2.10 incompatibility are not the same concrete defect." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "PyTorch-missing ImportError vs MacOS bus error loading a CLIP model; completely different failures." + "reason": "PyTorch-missing ImportError and a MacOS bus error with CLIP are completely different issues." } ] }, @@ -9568,8 +9520,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -9652,214 +9604,214 @@ "issue:28218|issue:32090", "issue:44206|issue:45823" ], - "prepared_review_unit_hash": "c034b1a8432e9bc54ca1c647d7268a11d3c4aa2f2d4e27d7404a2162d3eb8a2d", + "prepared_review_unit_hash": "99053abad02393e4ab786e8cbe6d34ddf14d2f0ce2d7f2dd02395efa3e24dc59", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36080, - "estimated_input_tokens": 9020, - "estimated_eval_tokens": 18296 + "serialized_chars": 36073, + "estimated_input_tokens": 9019, + "estimated_eval_tokens": 18294 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most soft links are false positives across unrelated model-specific regressions. The only clear duplicate pair is issue:45310 / issue:45478, which report the same Qwen3.5 MoE from_pretrained failure with identical wording.", - "confidence": 0.89, - "canonical_issue_reason": "issue:45310 is the best canonical issue because it is the earlier report of the exact-match duplicate pair with the same title and broader inbound context.", + "summary": "This cluster is mostly heterogeneous: it contains many unrelated bug reports, with only one clear duplicate pair. The only strong merge candidate is the Qwen3.5 MoE from_pretrained regression (45310/45478).", + "confidence": 0.93, + "canonical_issue_reason": "issue:45310 is the best canonical anchor because issue:45478 is a near-exact duplicate of the same Qwen3.5 MoE from_pretrained failure, and 45310 is the earlier report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the strongest representative issue in this set: it is a clean exact duplicate target, while the rest of the pairings are mostly unrelated near-matches.", + "best_issue_reason": "issue:45310 is the strongest issue candidate in this set because it has an exact duplicate (45478) and a clear, concrete failure mode; most other items are unrelated to each other.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Both are failures, but one is Flash Attention 2 test breakage and the other is a Trainer broadcast NoneType error; different bugs and code paths." + "reason": "Different failures: Flash Attention 2 tests vs a Trainer _gpu_broadcast_one NoneType error." }, { "left": "issue:43611", "right": "issue:44466", "accept": false, - "reason": "Both mention v5/loading behavior, but one is about base_model_prefix loading and the other about lm_head.weight serialization; not the same defect." + "reason": "Both involve loading/serialization regressions, but they affect different code paths and symptoms." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "Completely different model families and failures: SwitchTransformers sparse layer config vs BEiT legacy field loading." + "reason": "Unrelated models and bugs: SwitchTransformers sparse-layer config vs BEiT legacy-field loading." }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "Python parsing/import issue in DebertaV2 vs a qwen3_omni_moe helper behavior issue; unrelated code paths." + "reason": "Different modules and problems: DebertaV2 import parsing vs qwen3_omni_moe helper behavior." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "One is a torch.compile/SDPA failure for Bamba, the other is a CI test stale device override in GraniteSpeech; not duplicates." + "reason": "Different subsystems: torch.compile/SDPA model failure vs a CI test stale device override." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Different import/runtime environments and root causes: flash_attn lookup on Python 3.13 vs torch.distributed dependency on Windows ROCm." + "reason": "Different import/dependency failures on different platforms and libraries." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Both are v5 regressions, but one is Pegasus output drift and the other is recurrent state reset in modular_qwen3_5; different models and mechanisms." + "reason": "Different model behaviors: Pegasus v4/v5 output drift vs recurrent state reset in modular_qwen3_5." }, { "left": "issue:45290", "right": "issue:46326", "accept": false, - "reason": "Both involve chat templating, but one is tool-call/tokenize crash and the other is continue_final_message behavior for mistral_common tokenizers." + "reason": "Both touch chat templating, but one is an assistant-message/tool-call crash and the other is a mistral_common continue_final_message bug." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Different import errors for different symbols (_flash_supports_window_size vs HybridCache); no shared underlying bug." + "reason": "Both are import errors, but for different missing symbols and unrelated fixes." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both are apply_chat_template issues, but one crashes on batched padding=False input and the other returns all-zero assistant masks for multimodal inputs." + "reason": "Same broad API area, but one is a batching/padding crash and the other is incorrect assistant masks." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers sparse-layer creation is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "Different model/config logic; no shared underlying defect." }, { "left": "issue:43335", "right": "issue:45593", "accept": false, - "reason": "SwitchTransformers configuration bug and D-FINE auxiliary-loss bug are unrelated." + "reason": "Unrelated model bugs: SwitchTransformers config vs D-FINE auxiliary loss handling." }, { "left": "issue:43335", "right": "issue:44112", "accept": false, - "reason": "SwitchTransformers config and GraniteSpeech CI stale device override are unrelated issues." + "reason": "No shared code-path problem; one is model config creation, the other is a CI device-override test." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Embedding dtype mismatch during evaluate is unrelated to a pipeline translation-task error message regression." + "reason": "Different behaviors: dtype mismatch during evaluate vs a pipeline error message regression." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "Different subsystems and symptoms: LayoutLMv3 error-message clarity vs Trainer GPU broadcast NoneType failure." + "reason": "Different areas and symptoms; error-message clarity vs distributed broadcast type error." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Both are attention/offload related, but one is flash_attention.py s_aux handling and the other is Qwen3_5MoeTextRotaryEmbedding CPU offload incompatibility." + "reason": "Both involve offload/attention-related code, but the concrete failures are different and not the same bug." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "A trailing-slash import warning for local paths is unrelated to GLM-4.7 Flash model loading failures in tests." + "reason": "Different issues: trailing-slash import warning vs unit-test model loading failure." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Different failure modes: Llama 3.2 Vision loading regression vs output embedding reinitialization after resizing token embeddings." + "reason": "Different loading/weight-tie bugs affecting different models and mechanisms." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC from_latents/STE mismatch and unexpected position-id keys are unrelated bugs." + "reason": "Different symptoms and code paths: DAC latent consistency vs unexpected position-id keys." }, { "left": "issue:46431", "right": "issue:46566", "accept": false, - "reason": "One is missing chat template support, the other is weights left on meta with device_map='auto'; different loading paths." + "reason": "A missing chat template error is unrelated to DiffusionGemma meta-device loading." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "Documentation example errors for SigLIP2 and BeitImageProcessorFast label reduction are unrelated." + "reason": "Different model families and bugs: SigLIP2 docs example vs BEiT fast image processor label reduction." }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "BitNet packed-weight loading under accelerate and GraniteSpeech CI stale override are unrelated." + "reason": "BitNet packed-weight loading bug is unrelated to a GraniteSpeech CI test failure." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "CPU offload device mismatch for bark-small is unrelated to the LwDetrImageLoss AMP float16 crash." + "reason": "Both are runtime crashes, but one is CPU-offload device mismatch and the other is AMP/CUDA loss failure." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "Exact same title and same Qwen3.5 MoE from_pretrained regression; these are clear duplicates." + "reason": "Same title, same Qwen3.5 MoE from_pretrained regression, with 45478 clearly duplicating 45310." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Both are v4-v5 tokenizer/model regressions, but they affect different models and distinct behaviors." + "reason": "Both are regression reports, but they affect different models/tokenizers and different failure modes." }, { "left": "issue:39692", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 docs/example issues are unrelated to unexpected position-id key errors." + "reason": "No shared underlying bug; documentation/model mismatch is unrelated to position-id key warnings." }, { "left": "issue:26413", "right": "issue:31501", "accept": false, - "reason": "Resume-from-checkpoint epoch_iterator failure and FineWeb SLM training not starting are different training issues." + "reason": "Training startup/resume issues are too generic here; the concrete failure reports do not match." }, { "left": "issue:28218", "right": "issue:32090", "accept": false, - "reason": "Tokenizer added-space bug and GPU broadcast NoneType error are unrelated." + "reason": "Tokenizer spacing bug is unrelated to a Trainer distributed broadcast type error." }, { "left": "issue:44206", "right": "issue:45823", "accept": false, - "reason": "Different models and failures: LaserFeatureExtractor center-arg regression vs Gemma4 PLE device mismatch during forward." + "reason": "Different model/runtime failures with no shared concrete code path." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is well grounded in the packet. The duplicate pair issue:45310 / issue:45478 is an exact title match and the earlier issue is correctly chosen as canonical. The soft-edge rejects are appropriately conservative and the reasons generally track the issue titles without overstating evidence." + "feedback": "Grounded overall. The summary stays conservative, and the one accepted soft-edge pair (45310/45478) is supported by identical titles and chronology. The other verdicts are mostly cautious false negatives given the limited evidence in titles alone. No obvious overstatement or mergeability issue stands out." }, "evaluator_used": true, "retried": false, @@ -9875,8 +9827,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -9909,7 +9861,9 @@ "issue:44561", "issue:44617", "issue:44855", + "issue:44857", "issue:44977", + "issue:45072", "issue:45083", "issue:45127", "issue:45229", @@ -9923,10 +9877,8 @@ "issue:46192", "issue:46234", "issue:46399", - "issue:46421", "issue:46655", "issue:46697", - "issue:46858", "issue:47000" ], "soft_pairs": [ @@ -9936,6 +9888,7 @@ "issue:45459|issue:46234", "issue:32090|issue:33357", "issue:42490|issue:43575", + "issue:44857|issue:45072", "issue:45593|issue:47000", "issue:43550|issue:45083", "issue:38175|issue:44493", @@ -9956,199 +9909,204 @@ "issue:44977|issue:45229", "issue:43278|issue:43381", "issue:43927|issue:45698", - "issue:32090|issue:33290", - "issue:46421|issue:46858" + "issue:32090|issue:33290" ], - "prepared_review_unit_hash": "721f7b298fb8d315920a105884e95dacc29c608aa9ccc5119c0c6bad715a646b", + "prepared_review_unit_hash": "ff23fc159436158cf0ee3c3d9a76a655538301c7c497469c33903b931c3c40c8", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36274, - "estimated_input_tokens": 9069, - "estimated_eval_tokens": 18394 + "serialized_chars": 36206, + "estimated_input_tokens": 9052, + "estimated_eval_tokens": 18360 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a noisy set of mostly unrelated Transformers bug reports. The soft-similarity pairs are false positives: they share broad subsystems at most, but not the same concrete bug or fix path. No PRs are present.", - "confidence": 0.79, - "canonical_issue_reason": "issue:45920 is the strongest standalone report: it describes a broad tokenizer regression affecting multiple model families and has the highest signal for a representative issue.", + "summary": "These issues are broadly heterogeneous across models, processors, generation, loading, and training paths. I do not see any soft edge pair that looks like the same underlying bug/change, so all soft links should be rejected.", + "confidence": 0.96, + "canonical_issue_reason": "Issue:45920 is the strongest anchor: it reports a broad, concrete tokenizer regression affecting multiple models, with clear user impact and higher duplicate-likelihood than the more model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best cluster anchor overall because it is the most broadly impactful and well-supported issue in an otherwise heterogeneous set.", + "best_issue_reason": "Issue:45920 is the best overall representative because it is high-signal, broadly scoped, and plausibly central to nearby tokenizer-related reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "Different failure modes: LoRA/extended-vocab collapse vs category-count overflow; not the same bug." + "reason": "Different bug classes: LoRA/vocab merge collapse vs category-cardinality runtime error." }, { "left": "issue:44617", "right": "issue:46697", "accept": false, - "reason": "One is a CUDA OOM report, the other is an image padding bug in Swin2SR; unrelated code paths." + "reason": "Unrelated failures: video-model OOM vs image-processor over-padding." }, { "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "Checkpoint key mismatch and multimodal token-length helper behavior are distinct issues." + "reason": "Checkpoint key mismatch is unrelated to a multimodal feature-length helper bug." }, { "left": "issue:45459", "right": "issue:46234", "accept": false, - "reason": "Tokenizer protobuf error handling vs a missing import_utils availability symbol are separate bugs." + "reason": "Both touch import/utilities, but one is tokenizer error masking and the other is a missing import_utils symbol." }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Trainer broadcast of NoneType and MacOS bus error in CLIP are unrelated." + "reason": "Distributed broadcast NoneType error is not the same as a MacOS bus error on CLIP loading." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "Local-path import warning and tensor-parallel OOM are different problems." + "reason": "Local-path slash warning and tensor-parallel OOM are different load-path failures." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype/mixed precision, but they affect different code paths and symptoms." }, { "left": "issue:45593", "right": "issue:47000", "accept": false, - "reason": "Both are detection losses, but they are different models and different failure mechanisms; not mergeable as one fix." + "reason": "Different detection-loss bugs: missing auxiliary losses vs infeasible cost-matrix crash." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba is unrelated to qwen3_omni_moe feature-length arithmetic." + "reason": "torch.compile/SDPA failure is unrelated to a Qwen3 omni feature-length helper bug." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "Zero probabilities in SigLIP2 and unexpected position-id keys are distinct inference/model-loading issues." + "reason": "Zero probabilities in one model do not match a broad position-id key mismatch issue." }, { "left": "issue:27957", "right": "issue:34689", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta and Llama 3.2 Vision model loading are different code paths." + "reason": "Both are model-specific loading issues, but they involve different models and failures." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash-Attn test failures and embedding reinitialization after resize are unrelated." + "reason": "Flash Attention 2 test failures are unrelated to embedding reinitialization after resize." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat-template tool-call handling and Qwen2.5-VL video position IDs are separate multimodal bugs." + "reason": "Chat-template crashes with tool calls are unrelated to Qwen2.5-VL video position-id errors." }, { "left": "issue:29870", "right": "issue:32090", "accept": false, - "reason": "BART generate() decoder mask bug and distributed broadcast NoneType error are unrelated." + "reason": "Both are runtime errors, but one is a BART generate-path bug and the other is a trainer broadcast type error." }, { "left": "issue:10000", "right": "issue:15354", "accept": false, - "reason": "German DistilBertModel issue and torch.jit GeneratorExp export failure do not describe the same bug." + "reason": "Different frameworks and failures: DistilBert issue vs torch.jit.script GeneratorExp export error." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "Embedding dtype drift in eval and undefined video-branch helpers are different issues." + "reason": "Embedding dtype drift is not the same as undefined multimodal-token helper names in the video branch." }, { "left": "issue:44448", "right": "issue:44977", "accept": false, - "reason": "Pegasus v4/v5 output difference and Qwen3.5 flash-attention generation failure are unrelated." + "reason": "Sequence-to-sequence output regression vs flash-attention generation failure are distinct problems." }, { "left": "issue:45290", "right": "issue:46655", "accept": false, - "reason": "Chat-template crashes and video padding miscalculation are both multimodal, but not the same concrete defect." + "reason": "Chat-template tool-call handling is unrelated to video-processor padding calculation." }, { "left": "issue:44360", "right": "issue:44561", "accept": false, - "reason": "DSA indexer activation discussion and removal of is_torch_fx_available breaking trust_remote_code are unrelated." + "reason": "A missing ReLU in a discussion/implementation issue is unrelated to removal of a utility function breaking trust_remote_code models." }, { "left": "issue:44855", "right": "issue:46399", "accept": false, - "reason": "Python 3.13 torch.jit parsing/indentation failure and a missing decorator on rotary-pos-emb registration are not the same fix." + "reason": "Both mention decorators, but one is a Python 3.13 import/parser issue and the other is a registration/value-error in Qwen3.5 kernels." }, { "left": "issue:45920", "right": "issue:46192", "accept": false, - "reason": "Tokenizer ID regression across models and return_tensors being ignored are different tokenizer behaviors." + "reason": "Tokenizer ID regression is different from return_tensors being ignored when text_kwargs is passed." }, { "left": "issue:29870", "right": "issue:33290", "accept": false, - "reason": "A generation-time BART bug and a DeepSpeed Adafactor OOM are unrelated." + "reason": "Generate-path decoder_attention_mask crash is unrelated to deepspeed Adafactor OOM." }, { "left": "issue:37428", "right": "issue:42175", "accept": false, - "reason": "Missing import symbol and packaging/backend selection under pip extras are distinct issues." + "reason": "Flash-attention import failure and missing TensorFlow backend dependency are unrelated." }, { "left": "issue:40444", "right": "issue:43526", "accept": false, - "reason": "Qwen2.5-VL multi-image fine-tuning failure and BeitImageProcessorFast reduce_labels behavior are unrelated." + "reason": "Multi-image VL fine-tuning failure is not the same as BeitImageProcessorFast reduce_labels behavior." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation bug in Qwen3.5 and multi-GPU OOM in Gemma4 are different problems." + "reason": "Flash-attention generation bug and multi-GPU OOM are different failure modes." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "Eval dtype change and gradient-checkpointing-in-eval-mode are separate behaviors." + "reason": "Eval dtype mismatch is unrelated to gradient checkpointing being disallowed in eval mode." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Both involve save/load, but one loses custom token IDs and the other loads the wrong custom module; not the same bug." + "reason": "Both involve save/load behavior, but one is custom token IDs in config and the other is wrong custom-module resolution." }, { - "left": "issue:46421", - "right": "issue:46858", + "left": "issue:32090", + "right": "issue:33290", "accept": false, - "reason": "Lazy cache recompilation with chunked prefill and device-tensor seq_length arithmetic on non-CUDA backends are different cache issues." + "reason": "Trainer broadcast NoneType error is not the same as optimizer memory exhaustion." } ] }, @@ -10167,8 +10125,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -10186,7 +10144,6 @@ "issue:42490", "issue:42915", "issue:42994", - "issue:43065", "issue:43278", "issue:43381", "issue:43575", @@ -10196,7 +10153,6 @@ "issue:43756", "issue:43873", "issue:43976", - "issue:43994", "issue:44186", "issue:44206", "issue:44361", @@ -10219,9 +10175,12 @@ "issue:45588", "issue:45803", "issue:45823", - "issue:46121" + "issue:46121", + "issue:46421", + "issue:46858" ], "soft_pairs": [ + "issue:46421|issue:46858", "issue:43873|issue:45229", "issue:42490|issue:43873", "issue:43976|issue:45440", @@ -10248,205 +10207,210 @@ "issue:42915|issue:42994", "issue:43756|issue:45440", "issue:43653|issue:44442", - "issue:45372|issue:45588", - "issue:43065|issue:43994" + "issue:45372|issue:45588" ], - "prepared_review_unit_hash": "2f8f3431bcd11095f513da10e6f01dfb8b296bf0eb90c7f74d21d467d6d6dd6d", + "prepared_review_unit_hash": "00df70f7f1931bce5d3f779f609f340618c4ee9ec0c7d70c5184cdfa0d4b226a", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36198, - "estimated_input_tokens": 9050, - "estimated_eval_tokens": 18356 + "serialized_chars": 36287, + "estimated_input_tokens": 9072, + "estimated_eval_tokens": 18400 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: the suggested soft pairs mostly share broad ecosystem terms (quantization, generation, tokenizer, model loading) but not the same concrete bug or fix path. No pair looks safe to merge as a duplicate, and there is no true cluster-wide canonical duplicate.", - "confidence": 0.86, - "canonical_issue_reason": "Issue 43873 is the closest thing to a cluster anchor because it is relatively central/active, but it is not a real umbrella duplicate for the rest of these unrelated reports.", + "summary": "This cluster is mostly a noisy mix of unrelated Transformers issues; the soft links are theme-level similarities, not true duplicates. No soft edge looks like the same underlying bug/change.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43873 is the closest thing to a hub because it has one of the few multiple soft links and sits in the broader quantization/offloading area, but even that match is weak and the cluster is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "43873 is the best issue only as a triage anchor; it has the strongest centrality, but the surrounding issues do not describe the same defect.", + "best_issue_reason": "issue:43873 is the best representative only in a loose thematic sense; it is not a strong duplicate canonical because the linked issues are mostly about different failures.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache/prefill/compile behavior, but one is about lazy cache init causing recompilation while the other is a device-tensor seq-length bug in static cache arithmetic." + }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Both involve memory/offloading themes, but one is quantization offloading behavior and the other is Gemma4 multi-GPU OOM; different symptom and code path." + "reason": "Quantization/offloading vs multi-GPU CUDA OOM are different failure modes and likely different root causes." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Local-path import warning vs quantization/offloading failure; unrelated behaviors." + "reason": "Trailing-slash local model import warning is unrelated to quantization offloading behavior." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python/version compatibility issue vs DeepSeekV3 implementation divergence; not the same defect." + "reason": "Python version compatibility issue and DeepSeekV3MoE divergence are unrelated." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "Both are v5 regressions, but one is Pegasus generation output drift and the other is lm_head serialization/tied-weights behavior." + "reason": "Both are v5 regressions, but one is output drift for Pegasus and the other is serialization of tied weights; different code paths." }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "Pipeline top_k nesting bug vs trainer broadcast NoneType error; entirely different code paths." + "reason": "Top-k pipeline shape change and Trainer broadcast NoneType are unrelated bugs." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Tokenizer special-token handling vs Gemma3 NaN embeddings with sliding-window attention; unrelated." + "reason": "Tokenizer special-token handling and Gemma3 mixed-length NaN embeddings are different subsystems and symptoms." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma eos_token_id config mismatch vs Qwen3.5 flash-attention generation failure; different models and failure modes." + "reason": "Gemma eos_token_id config mismatch and Qwen3.5 flash-attention generation failure are not the same bug." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error-message clarity vs TrainerState token counting not updating; not the same bug." + "reason": "LayoutLMv3 error-message clarity and TrainerState token-count update are unrelated." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "Removal of a utility breaking trust_remote_code vs Deepseek tokenizer regression; separate breakages." + "reason": "Removal of is_torch_fx_available breaking remote-code models is distinct from Deepseek tokenizer correctness." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Eval-mode gradient checkpointing restriction vs resume_from_checkpoint key mismatch; unrelated." + "reason": "Checkpointing in eval mode and resume_from_checkpoint key mismatch are different training-state issues." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS bus error on CLIP model loading vs SigLIP2 documentation example issues; not the same underlying bug." + "reason": "MacOS bus error on CLIP loading and SigLIP2 doc example errors are separate problems." }, { "left": "issue:45588", "right": "issue:45823", "accept": false, - "reason": "Flash-attention s_aux=None crash vs Gemma4 device-map mismatch; both Gemma4-adjacent but distinct code paths." + "reason": "Flash-attention s_aux=None crash and Gemma4 PLE device mismatch are different forward-path bugs." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Wav2Vec2 save/tokenization failures vs a bare-except bug in Fuyu tensor conversion; unrelated." + "reason": "Wav2Vec2 save/tokenization failure and Fuyu bare-except swallowing errors are unrelated." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "Training/evaluate dtype mismatch vs tensor-parallel OOM on Qwen2-57B; different issues." + "reason": "Embedding dtype drift in eval and tensor-parallel OOM on Qwen2-57B are not the same underlying issue." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Qwen3 auto device mapping cuda assert vs Gemma4 multi-GPU OOM; both are GPU issues but not the same bug." + "reason": "Both involve large-model multi-GPU/device mapping, but one is a cudaErrorAssert on Qwen3 and the other is OOM on Gemma4." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Qwen3.5 flash-attention generation regression vs stale num_return_sequences warning in paged generate; unrelated." + "reason": "Flash-attention generation failure and stale num_return_sequences warning are unrelated." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "LayoutLMv2 tokenizer crash vs Wav2Vec2 save/tokenization failure; different tokenizers and failure modes." + "reason": "LayoutLMv2 tokenizer crashes and Wav2Vec2 save/tokenization failure are different tokenizer/data-path bugs." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "base_model_prefix loading breakage vs rope-parameter dict conversion TypeError; unrelated." + "reason": "base_model_prefix loading regression and rope-parameter dict conversion TypeError are unrelated." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "MLukeTokenizer attribute error vs Wav2Vec2 save/tokenization failures; not the same defect." + "reason": "MLuke tokenizer AttributeError and Wav2Vec2 save/tokenization failure do not share a concrete root cause." }, { "left": "issue:29870", "right": "issue:34689", "accept": false, - "reason": "BART generate unbound variable vs Llama 3.2 vision model-loading breakage; unrelated." + "reason": "BART generate decoder_attention_mask bug and Llama 3.2 vision loading breakage are different model-specific regressions." }, { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "BigBirdTokenizer special-token registration vs Qwen3_5 recurrent-state reset with cache; different subsystems." + "reason": "BigBirdTokenizer special-token registration and Qwen3_5 recurrent-state reset are unrelated." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Laser feature extractor center-arg regression vs Qwen multimodal video-input regression; unrelated." + "reason": "LasrFeatureExtractor center-arg crash and Qwen video-input regression are different code paths." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention 2 test failures vs image processor void-segmentation-map handling; both test regressions but not the same bug." + "reason": "Flash Attention 2 test failures and void segmentation-map processing are only both test/vision-related at a high level, not the same bug." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "Qwen3Moe FP8 config failure vs quantized model saving failure; both quantization-related but different operations." + "reason": "Both involve quantization, but one is FineGrainedFP8Config load failure and the other is quantized model saving; different operations." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch vs DeepSeekV3MoE divergence; different models and implementations." + "reason": "Smollm3 RoPE-layer mismatch and DeepSeekV3MoE divergence are unrelated architecture issues." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "Tokenizer special-token bug vs AutoTokenizer failing to load a speech tokenizer; unrelated." + "reason": "BigBirdTokenizer mask-token bug and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer failures." }, { "left": "issue:45372", "right": "issue:45588", "accept": false, - "reason": "Gemma4 processor import error vs flash-attention s_aux=None crash; same model family, different failure layers." + "reason": "Gemma 4 processor import failure and flash-attention s_aux=None crash are unrelated." }, { - "left": "issue:43065", - "right": "issue:43994", + "left": "issue:46421", + "right": "issue:46858", "accept": false, - "reason": "Sam3 pixel-decoder dummy Conv2d vs SigLIP2 nonsensical outputs with AutoModel/pipeline; unrelated." + "reason": "Both mention cache behavior, but the first is about recompilation from lazy init while the second is about device-tensor seq-length breaking slice arithmetic on non-CUDA backends." } ] }, @@ -10465,8 +10429,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -10519,6 +10483,7 @@ "issue:46701" ], "soft_pairs": [ + "issue:43065|issue:43994", "issue:43065|issue:45412", "issue:43412|issue:45588", "issue:44265|issue:45198", @@ -10553,232 +10518,226 @@ "issue:43526|issue:44186", "issue:45250|issue:46640" ], - "prepared_review_unit_hash": "91bf480a175a9a7d9c0e7281ba90f02a60e942a624a931cec22cf9cec159c0be", + "prepared_review_unit_hash": "983fe489bfd50d880e21e3148c77cc512524af79559e2fba2faf4bb9eddb8e91", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 36788, - "estimated_input_tokens": 9197, - "estimated_eval_tokens": 18650 + "soft_pair_count": 34, + "serialized_chars": 37089, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 }, "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 bugs, and the soft pairs look like similarity false positives rather than true duplicates. There is no strong single duplicate set, but the Flash Attention 2 performance-regression issue is the closest thing to a representative anchor.", - "confidence": 0.84, - "canonical_issue_reason": "If one issue must anchor the cluster, #46693 is the most concrete and discussion-heavy FA2-related bug; the rest of the items mostly do not describe the same underlying problem.", + "summary": "This cluster is a broad mix of unrelated Transformers issues across model loading, tokenizers, quantization, export, memory, and attention. None of the soft pairs look like the same underlying bug with enough specificity to merge.", + "confidence": 0.92, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "#46693 is the most actionable representative among the mostly disconnected entries because it states a specific failure mode and code-path context, unlike the very generic FA2 titles.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Different models and symptoms: Sam3PixelDecoder dummy Conv2d vs Siglip2 nonsensical outputs. No shared code-path evidence." + }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Different problems: a dummy Conv2d placeholder in Sam3PixelDecoder vs an RT-DETR memory-leak/deallocation issue." + "reason": "Sam3 pixel-decoder placeholder layer and RT-DETR GC memory leak are unrelated bugs." }, { "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Different code paths: executorch export guards/erfinv support vs flash_attention handling of s_aux=None." + "reason": "Gemma3n export failure and flash_attention.py s_aux=None crash are different failure modes in different paths." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "Unrelated failures: torch.export with torch_compilable_check vs Wav2Vec2 save_pretrained/tokenization." + "reason": "torch.export with torch_compilable_check is unrelated to Wav2Vec2 save_pretrained/tokenization failure." }, { "left": "issue:45491", "right": "issue:46693", "accept": false, - "reason": "Both mention FA2/attention, but one is a Gemma3 NaN correctness bug and the other is a long-context performance regression." + "reason": "Both mention FA2, but one is NaN embeddings from sliding-window padding and the other is a performance regression in long-context generation." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "Completely different models and symptoms: LayoutLMv3 error messaging vs a MacOS bus error with CLIP." + "reason": "LayoutLMv3 box validation messaging and CLIP bus error on macOS have no common underlying bug." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "OOM on loading a large Qwen model is not the same issue as resume_from_checkpoint key mismatch." + "reason": "Tensor-parallel OOM on Qwen2-57B is unrelated to resume_from_checkpoint key mismatch." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "Category cardinality RuntimeError is unrelated to a stale generate() warning." + "reason": "Category-count limit error and stale num_return_sequences warning are different issues in different code paths." }, { "left": "issue:45459", "right": "issue:45712", "accept": false, - "reason": "Tokenizer error masking with missing protobuf is unrelated to leftover dummy classes leaking into dir(transformers)." + "reason": "Tokenizer error masking when protobuf is missing is unrelated to leftover dummy classes leaking without torch." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Local-path import warning and gradient-checkpointing-in-eval are different behaviors in different subsystems." + "reason": "Local-path import warning and gradient-checkpointing-in-eval behavior are unrelated." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "Error-message clarity for LayoutLMv3 boxes does not match token-embedding reinit after resize." + "reason": "Error-message clarity for LayoutLMv3 boxes is unrelated to embedding reinitialization after resize." }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "DeepseekV3MoE implementation divergence is unrelated to D-FINE auxiliary-loss behavior." + "reason": "DeepseekV3MoE parity divergence and D-FINE auxiliary-loss omission are different model-specific bugs." }, { "left": "issue:45372", "right": "issue:45406", "accept": false, - "reason": "Both involve Gemma4, but one is an import dependency failure and the other is a missing attribute in serve." + "reason": "Importing ReasoningEffort from mistral_common and Gemma4Processor missing _tokenizer in serve are distinct problems." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "Siglip2 inference nonsense and RT-DETR memory release are unrelated bugs." + "reason": "Siglip2 inference quality issue and RT-DETR memory leak are unrelated." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Runtime special-token post-processor updates are unrelated to Gemma3 sliding-window NaNs." + "reason": "Runtime special-token post-processor update bug is unrelated to Gemma3 sliding-window NaNs." }, { "left": "issue:45406", "right": "issue:46701", "accept": false, - "reason": "Gemma4 processor serve crash and DiffusionGemma image-text pipeline failure are different concrete issues." + "reason": "Gemma4 serve attribute error and DiffusionGemma image-text-to-text pipeline failure are different processors/code paths." }, { "left": "issue:45412", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but one is a leak on deletion and the other is an overflow when calling sam3-litetext." + "reason": "RT-DETR GC memory leak and sam3-litetext memory overflow are not the same issue." }, { "left": "issue:43066", "right": "issue:44462", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as ignoring tokenizer.json during AutoTokenizer loading." + "reason": "Wrong decoder type in Transformers v5 and AutoTokenizer ignoring tokenizer.json are separate tokenizer-loading bugs." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "Gemma4 import failure from mistral_common is unrelated to protobuf-error masking in tokenizer code." + "reason": "Dependency import failure for Gemma4 and protobuf-masked tokenizer errors are unrelated." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Large-model TP OOM and int4 quantization reserved-memory regression are different memory problems." + "reason": "Both are OOM-related, but one is TP load OOM for Qwen2-57B and the other is increased CUDA reserved memory under int4 quantization." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "FastSpeech2ConformerTokenizer loading failure is unrelated to dtype mismatches in bfloat16 inference." + "reason": "FastSpeech2ConformerTokenizer loading and bfloat16 dtype mismatches are unrelated." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is unrelated to Whisper empty-transcription handling after align_special_tokens." + "reason": "Compiled-forward chunked generation inconsistency and Whisper empty-transcription failure after align_special_tokens do not share a code-path." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug and Kosmos2.5 OCR index error are different image-processing failures." + "reason": "BeitImageProcessorFast label reduction bug and Kosmos2.5 long-OCR index error are different preprocessing issues." }, { "left": "issue:43526", "right": "issue:46693", "accept": false, - "reason": "reduce_labels behavior is unrelated to FA2 long-context performance regression." + "reason": "No meaningful overlap: image-label reduction bug vs FA2 long-context generation regression." }, { "left": "issue:27957", "right": "issue:33290", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Adafactor/deepspeed OOM are unrelated." + "reason": "XLMRoberta Flash Attention 2 support and adafactor/deepspeed OOM are unrelated." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "Missing Qwen2_5_VL import symbol is not the same as a generic dependency issue." + "reason": "Specific Qwen2_5_VL import error is not the same as a generic dependency issue report." }, { "left": "issue:27957", "right": "issue:29127", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta is unrelated to LayoutLMv3 error messaging." - }, - { - "left": "issue:43994", - "right": "issue:46369", - "accept": false, - "reason": "Siglip2 inference errors do not match sam3-litetext memory overflow." - }, - { - "left": "issue:42915", - "right": "issue:43931", - "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe is unrelated to Qwen3-VL weight-shape mismatch." + "reason": "Flash Attention support for XLMRoberta and LayoutLMv3 box error messaging are unrelated." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "Adafactor/deepspeed OOM and embedding reinitialization after resize are different bugs." + "reason": "Optimizer/deepspeed OOM and embedding reinitialization after resize are different bugs." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "These are separate issues: optimizer OOM vs model-loading breakage for Llama 3.2 Vision." + "reason": "Adafactor/deepspeed OOM and Llama 3.2 90B Vision loading breakage are not the same issue." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "AMP/CUDA crash in LwDetrImageLoss is unrelated to flash_attention.py crashing on s_aux=None." + "reason": "LwDetr float16 AMP crash and flash_attention.py s_aux=None crash are distinct failures." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "BeitImageProcessorFast label handling and LayoutLMv2Tokenizer padding/NER crashes are different components." + "reason": "BeitImageProcessorFast reduce_labels bug and LayoutLMv2Tokenizer NER/padding crash are unrelated." }, { "left": "issue:45250", "right": "issue:46640", "accept": false, - "reason": "Both are generic FA2 titles, but there is no evidence they describe the same concrete bug or code path." + "reason": "Both are generic FA2 titles, but there is no evidence they describe the same concrete bug or change." } ] }, @@ -10797,8 +10756,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -10878,190 +10837,190 @@ "issue:43316|issue:45759", "issue:43749|issue:43931" ], - "prepared_review_unit_hash": "aa53702bc564a29006aec3be94490b3d81f1964ca564666f85c5e53a8ed06d00", + "prepared_review_unit_hash": "9bec96e22763cdbea3c2219448300edd6e94a1d967bad0d130f7fba4b5a2286e", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35128, - "estimated_input_tokens": 8782, - "estimated_eval_tokens": 17820 + "serialized_chars": 35122, + "estimated_input_tokens": 8781, + "estimated_eval_tokens": 17818 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of superficially similar but distinct transformer bug reports. The soft links are driven by shared terms like config, loading, offload, vision/video, and model families, but the underlying code paths and fixes differ enough that none should be merged as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "issue:43873 is the broadest and most central item here: it is open, has the most discussion/inbound references, and concerns a widely impactful offloading/quantization regression rather than a narrow model-specific edge case.", + "summary": "The set is a heterogeneous mix of unrelated Transformer bugs. None of the soft pairs look like the same underlying defect; most only share a broad subsystem or model family name, which is not enough to merge them as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43873 is the strongest standalone anchor in the set because it has the most discussion and inbound references, but the overall collection is too diverse to form a true duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest representative issue for the cluster because it has the highest engagement and the most general user impact.", + "best_issue_reason": "Issue 43873 is the most substantively triaged and highest-activity issue here, so it is the best representative if a single issue must be chosen.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Different failures: missing pad_token_id on Qwen3-VL config vs sliding_window behavior in Qwen3-MoE." + "reason": "Both are Qwen-related, but one is a missing pad_token_id config error and the other is a sliding_window behavior issue; different code paths and symptoms." }, { "left": "issue:46121", "right": "issue:46399", "accept": false, - "reason": "Unrelated regressions: rope validation TypeError vs missing decorator for apply_rotary_pos_emb registration." + "reason": "Rope validation TypeError and Qwen3.5 kernel registration failure are unrelated bugs despite both touching rotary/rope-adjacent code." }, { "left": "issue:43381", "right": "issue:43873", "accept": false, - "reason": "Different code paths: gradient checkpointing in eval mode vs offloading with quantization." + "reason": "Gradient checkpointing in eval mode and quantization offloading are separate behaviors with no concrete shared failure mode." }, { "left": "issue:45588", "right": "issue:46821", "accept": false, - "reason": "One is a flash_attention sink-less crash on s_aux=None; the other is a TimesFM window_size attribute error." + "reason": "Flash-attention sink handling and TimesFM window_size AttributeError are unrelated." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Both involve Qwen2.5-VL, but one is batched apply_chat_template padding handling and the other is video position ids/rope indexing." + "reason": "Both involve Qwen2.5-VL, but one is batched chat-template padding and the other is video vision_position_ids; not the same bug." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "Transformers 5.0 custom model init breakage is broader and different from patchtsmixer post_init policy." + "reason": "Custom model init regression and patchtsmixer post_init validation are different initialization issues, not one mergeable fix." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and category-count RuntimeError are unrelated." + "reason": "Qwen3.5 flash-attention generation and category cardinality limit errors are unrelated." }, { "left": "issue:43761", "right": "issue:45859", "accept": false, - "reason": "Different model families and symptoms: CLIPVision hidden_states regression vs Qwen3_5Moe rotary embedding CPU offload incompatibility." + "reason": "CLIPVisionModel hidden_states regression and Qwen3.5Moe CPU offload incompatibility affect different models and mechanisms." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Qwen3.5 recurrent-state reset with cache is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Recurrent cache reset in Qwen3.5 modular code and Whisper batch_decode skipping special tokens are unrelated." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Audio processor sample-rate default mismatch is unrelated to LasrFeatureExtractor passing an unsupported center arg." + "reason": "Audio chat-template sampling-rate defaulting and LASR feature extractor center-arg crash are different audio-processing bugs." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "Different tokenizer/model loading problem for FastSpeech2ConformerTokenizer vs Perceiver interpolation failure on non-default resolution." + "reason": "Tokenizer load failure for FastSpeech2Conformer and Perceiver image classification padding/resizing are unrelated." }, { "left": "issue:42490", "right": "issue:44743", "accept": false, - "reason": "Import warning for trailing-slash local model paths is unrelated to Qwen3.5 recurrent state handling." + "reason": "Local path import warnings and Qwen3.5 recurrent-state reset are not the same underlying issue." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Packaging/backend dependency issue and missing import_utils helper are not the same bug." + "reason": "Torch install backend packaging and missing import_utils availability helper are unrelated packaging/API issues." }, { "left": "issue:45003", "right": "issue:45440", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to DeepseekV3MoE divergence from remote implementation." + "reason": "Unsafe sys.modules access and DeepseekV3MoE divergence are distinct implementation problems." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to tokenizer loading failure." + "reason": "BitNet packed-weight loading and tokenizer loading for FastSpeech2Conformer are unrelated." }, { "left": "issue:27957", "right": "issue:29870", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support is unrelated to BART generate() decoder_attention_mask assignment error." + "reason": "XLMRoberta Flash Attention 2 support and BART generate() decoder_attention_mask error are unrelated model-specific bugs." }, { "left": "issue:43316", "right": "issue:43525", "accept": false, - "reason": "Gemma3TextConfig API discrepancy and Llama4Config missing pad_token_id are separate config-model mismatches." + "reason": "Both are config/API shape complaints, but Gemma3TextConfig inconsistency and Llama4Config missing pad_token_id are different concrete failures." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "patchtsmixer post_init policy and Qwen video-input regression are different model-specific issues." + "reason": "patchtsmixer post_init deprecation and Qwen video-input regression are unrelated." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation discrepancy and docs still referencing pipeline() are unrelated." + "reason": "SigLIP2 implementation drift and docs still referencing pipeline() are completely separate concerns." }, { "left": "issue:44062", "right": "issue:45907", "accept": false, - "reason": "AddedToken special-argument collision is unrelated to list_repo_files revision forwarding." + "reason": "AddedToken keyword duplication and list_repo_files revision forwarding are unrelated API bugs." }, { "left": "issue:45325", "right": "issue:46697", "accept": false, - "reason": "Qwen2.5-VL rope scaling bug and Swin2SR over-padding are different preprocessing/model bugs." + "reason": "Qwen2.5-VL rope position scaling and Swin2SR image over-padding are unrelated image/video preprocessing issues." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading regression does not match Gemma-3 eos_token_id config inconsistency." + "reason": "Quantization offloading failure and Gemma-3 eos_token_id inconsistency do not share a concrete code-path bug." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Torch.jit parser IndentationError on Python 3.13 is unrelated to DeepSpeed ZeRO3 deque pop error." + "reason": "Python 3.13 torch.jit parsing error and DeepSpeed ZeRO3 deque pop error are unrelated." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "Missing fields from trust_remote_code AutoConfig and GPTNeoX rotary_pct reload regression are distinct config serialization issues." + "reason": "trust_remote_code missing fields and GPTNeoXConfig rotary_pct reset are both config-related, but they are different persistence/loading bugs." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Gemma3TextConfig API discrepancy is unrelated to AutoModelForCausalLM.from_config not unwrapping text_config." + "reason": "Composite-config API discrepancy and AutoModelForCausalLM.from_config text_config unwrapping are related only at a high level, not the same bug." }, { "left": "issue:43749", "right": "issue:43931", "accept": false, - "reason": "FSDP CPU RAM efficient loading breakage is a different loading path than Qwen3-VL weight shape mismatch." + "reason": "FSDP CPU-efficient loading breakage and Qwen3-VL weight-shape mismatch are separate loading failures." } ] }, @@ -11080,8 +11039,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -11165,20 +11124,20 @@ "issue:43452|issue:44496", "issue:45803|issue:46365" ], - "prepared_review_unit_hash": "74470f264836aa2d03c32d8ceda8bf0af05b206df288c2d75ab7be30c865915a", + "prepared_review_unit_hash": "0d368db69fa49ab5fb8fbbfbd3d4780d0f40bf1d254ce9bd02aac909a2fd9914", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36304, + "serialized_chars": 36303, "estimated_input_tokens": 9076, "estimated_eval_tokens": 18408 }, @@ -11187,196 +11146,193 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous cluster, but the ModelOutput/output-field bug class appears in more than one issue: #44079 is the generic key-assignment report, and #43761 plus #46829 are concrete downstream misassignment/field-confusion symptoms. So this is better described as one small output-container subcluster plus many unrelated reports, not a one-off singleton.", - "confidence": 0.71, - "canonical_issue_reason": "#44079 is still the clearest canonical issue because it states the underlying generic ModelOutput key-assignment defect directly, while #43761 and #46829 read as model-specific symptoms of that same output-field confusion.", + "summary": "The candidate edges are mostly false positives: they share broad Transformers themes, but each pair points to a different concrete failure mode, model, or code path. No soft edge is safe to merge as a duplicate.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45698 is the broadest and most discussed save/load regression in the set, so it is the least-bad anchor issue; however, the overall cluster is heterogeneous and not a true duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "#44079 is the best cluster representative because it captures the broad root-cause bug class rather than a single model\u2019s manifestation.", + "best_issue_reason": "Issue 45698 is the best representative issue by scope and discussion activity, but only as a loose anchor\u2014not as a duplicate canonical for the rest of the items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44987", "right": "issue:46178", "accept": false, - "reason": "Different failure modes: a transformers>=5.1.0 loading regression vs a missing tensorflow_text backend mapping." + "reason": "Different bugs: model-loading failure on a specific repo/version vs a missing backend registration for tensorflow_text." }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Unrelated paths: local dataset loading in a training script vs model config recognition/loading." + "reason": "Different failure classes: local dataset loading in a training script vs unrecognized model/config missing model_type." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation state handling after sync_gpus is not the same bug as gather_for_metrics truncating tuple labels." + "reason": "Generation kwargs sync bug after sync_gpus vs gather_for_metrics dropping tuple labels in the last batch." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer loading failure and a bare-except in Fuyu tensor conversion are separate issues." + "reason": "Tokenizer import/load failure vs a bare except swallowing interrupts and masking errors; unrelated code paths." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "Both involve Qwen/VL model loading/saving, but one is a weight-shape mismatch and the other is a save_pretrained regression." + "reason": "One is a weight-shape mismatch when loading, the other is a save_pretrained regression for Qwen3.5 checkpoints." }, { "left": "issue:43761", "right": "issue:44079", - "accept": true, - "reason": "Yes: CLIPVisionModel returning hidden_states=None fits the same ModelOutput key-assignment bug class described by #44079." + "accept": false, + "reason": "Hidden-states output regression in CLIPVisionModel vs ModelOutput key assignment logic; different layers of the stack." }, { "left": "issue:43701", "right": "issue:46506", "accept": false, - "reason": "Checkpoint resume key mismatch and CPU/CUDA result divergence are different problems." + "reason": "Checkpoint resume key mismatch vs CPU/CUDA numerical divergence; not the same underlying bug." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "ModelOutput assignment is unrelated to torch.compile attention branching." + "reason": "CLIP hidden_states output bug vs attention-kernel divergence under torch.compile; different symptoms and fixes." }, { "left": "issue:46829", "right": "issue:46932", "accept": false, - "reason": "VideoPrism output-field confusion and Gemma4 quantized embedder dtype casting are unrelated." + "reason": "VideoPrism output-field bug vs Gemma4 quantized embedder dtype casting; different models and code paths." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py launch failure and tensor-parallel OOM are different distributed-runtime issues." + "reason": "Generic 3D parallel script failure vs OOM when loading a specific Qwen2 model with tensor parallelism." }, { "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "torch.compile + SDPA behavior is unrelated to resume_from_checkpoint key mismatches." + "reason": "torch.compile/SDPA failure in Bamba vs resume_from_checkpoint key mismatch; unrelated runtime behaviors." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "Tensor-parallel OOM and config/key loading errors are distinct load-time failures." + "reason": "OOM during tp load of a large model vs key error loading MobileLLM-125M; different causes." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 support and NoneType broadcast errors are unrelated." + "reason": "Flash Attention 2 support issue for XLMRoberta vs Trainer GPU broadcast NoneType error; no shared bug." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch and checkpoint key mismatch are separate failures." + "reason": "CPU offload device mismatch vs checkpoint resume key mismatch; different subsystems entirely." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Pipeline top_k nesting behavior is unrelated to missing PyTorch in the environment." + "reason": "Pipeline top_k output nesting bug vs missing PyTorch import error; unrelated." }, { "left": "issue:42907", "right": "issue:45698", "accept": false, - "reason": "Both are serialization-adjacent, but dequantized save/load breakage is not the same bug as wrong custom-module loading." + "reason": "Both involve save/load behavior, but one is quantized-model saving after dequantization and the other is custom-module restoration; not the same bug." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Processor/tokenizer regression and a misleading pipeline error message are different issues." + "reason": "Processor/tokenizer regression with images vs an error-message wording bug for unsupported translation tasks." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "A model_max_length question is unrelated to the transformers.onnx import error." + "reason": "Question about model_max_length vs missing transformers.onnx module import; no duplicate relationship." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "Custom processor/tokenizer regression and Gemma eos_token_id inconsistency are different regressions." + "reason": "Different issues: custom model code breaking around processor.tokenizer/images vs eos_token_id config inconsistency." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Gemma eos_token configuration mismatch is unrelated to Whisper empty-transcription generation failures." + "reason": "Gemma eos_token_id mismatch vs Whisper empty-transcription generation failure after align_special_tokens; different models and failure points." }, { "left": "issue:46234", "right": "issue:46291", "accept": false, - "reason": "Missing import_utils helper and kernels v0.15 version/revision requirements are different dependency issues." + "reason": "Missing import_utils helper vs pipelines import requiring kernels version/revision; unrelated dependency issues." }, { "left": "issue:44464", "right": "issue:46421", "accept": false, - "reason": "Both mention chunked generation, but incorrect outputs and spurious recompilation are not the same bug." + "reason": "Chunked generation output inconsistency under compiled forward vs spurious recompilation from lazy cache init; same area but not the same bug." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Gemma3TextConfig API discrepancy and Mistral-3 dtype mismatch are unrelated." + "reason": "Gemma3TextConfig API mismatch vs Mistral-3 image-preprocessor dtype mismatch; different model families and fixes." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "Optional post_init policy and non-persistent buffer junk in v5 are separate release regressions." + "reason": "Optional post_init policy for patchtsmixer vs non-persistent buffer junk in v5; unrelated issues." }, { "left": "issue:43316", "right": "issue:46821", "accept": false, - "reason": "Gemma3TextConfig API mismatch is unrelated to TimesFM window_size AttributeError." + "reason": "Gemma3TextConfig discrepancy vs TimesFM window_size AttributeError; different APIs and models." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "Generic custom model initialization breakage and GPTNeoX rotary_pct reload reversion are different configuration bugs." + "reason": "Custom model initialization breakage in Transformers 5 vs GPTNeoX rotary_pct not persisting on reload; different bugs." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "Gemma eos_token_id mismatch is unrelated to sinq quantized save/load breakage." + "reason": "eos_token_id configuration mismatch vs quantized sinq save_pretrained/from_pretrained breakage; both serialization-adjacent but not duplicates." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "gguf_file loading issues are different from an unrecognized config missing model_type." + "reason": "gguf_file loading bug for tokenizer/model vs missing model_type in config.json; separate loading failures." }, { "left": "issue:45803", "right": "issue:46365", "accept": false, - "reason": "Fuyu's bare except swallowing errors is unrelated to Trainer import failure triggered by datasets." + "reason": "Bare except in Fuyu tensor conversion vs Trainer import failing until datasets is uninstalled; unrelated import/runtime problems." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The main conclusion is a bit too strong. While the 43761\u219444079 pairing is plausible from the titles, the summary says this is the *only* issue that shares the ModelOutput bug class; issue:46829 also looks like the same kind of output-field misassignment problem (`hidden_states`/`last_hidden_state` confusion). So the cluster is not quite as clearly one-off/heterogeneous as summarized, and the evidence supports a more cautious statement." - }, - "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 }, @@ -11388,8 +11344,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -11469,20 +11425,20 @@ "issue:44734|issue:45406", "issue:44855|issue:45588" ], - "prepared_review_unit_hash": "4ab173e85904a4c9fe837933de0f7f2814e45a11f5c9d600ae8b16e2b8fc52f3", + "prepared_review_unit_hash": "258729f45e4d05d1b91b42e942e994b72f1d846fa82d8b9110febf26b0d0a190", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 25, - "serialized_chars": 35156, + "serialized_chars": 35154, "estimated_input_tokens": 8789, "estimated_eval_tokens": 17834 }, @@ -11491,162 +11447,162 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No single duplicate cluster emerges here: the soft pairs are mostly broad textual matches across unrelated models, processors, cache paths, and training bugs. All soft edges should be rejected.", - "confidence": 0.96, - "canonical_issue_reason": "No clear canonical issue: the items span many unrelated bugs and fixes, so treating any one as the cluster representative would be misleading.", - "canonical_pr_reason": "No PRs are present in the provided cluster.", - "best_issue_reason": "No issue is a strong global representative because the cluster is not one underlying problem; if forced, any choice would be arbitrary.", - "best_pr_reason": "No PRs are present in the provided cluster.", + "summary": "The set is heterogeneous: most items are distinct model-, processor-, or config-specific bugs with only superficial wording overlap. I found no soft-edge pair that clearly describes the same underlying defect or change, so all candidate merges are rejected.", + "confidence": 0.91, + "canonical_issue_reason": "issue:44534 is the broadest cross-cutting regression in the set and most suitable as a representative issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44534 has the widest impact and a general Transformers v5 failure mode, making it the best single issue representative here.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44162", "right": "issue:44792", "accept": false, - "reason": "ESM2 model breakage and a Janus image-generation test failure are unrelated bugs with different code paths." + "reason": "Unrelated domains: ESM2 model breakage vs Janus image-generation test failure." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "One is a tied lm_head weight bug in Mistral3; the other is a Flash Attention support error in Ministral-3." + "reason": "Both concern Mistral/Ministral models, but one is a tied-weight bug and the other is a Flash Attention support failure." }, { "left": "issue:44242", "right": "issue:46894", "accept": false, - "reason": "Load-balancing loss gating and ASR training-loss double-shift are different loss-calculation bugs." + "reason": "Different loss bugs in different training setups; only a generic 'training loss' theme is shared." }, { "left": "issue:45632", "right": "issue:46424", "accept": false, - "reason": "A local trust_remote_code cache-path collision is unrelated to max_cache_length handling in quantized cache." + "reason": "One is a trust_remote_code cache-path collision, the other is cache-length handling for quantized cache." }, { "left": "issue:44610", "right": "issue:46369", "accept": false, - "reason": "OmDet-Turbo image-size mismatch and SAM3 LiteText memory overflow are different model-specific problems." + "reason": "Different SAM3 problems: processor/model image-size mismatch vs memory overflow." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "These are different multimodal processor crashes with different inputs and failure points." + "reason": "Qwen3OmniMoe video-input processing vs Qwen2.5-VL batched chat templating; not the same bug." }, { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "A dummy Conv2d placeholder in Sam3PixelDecoder is not the same issue as a video OOM in Sam3Video." + "reason": "Both mention SAM3 video/pixel decoder, but one is a dummy layer issue and the other is OOM behavior." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "TensorBoard/TensorFlow backend install handling and a missing import_utils symbol are unrelated." + "reason": "Completely different backend/import issues with no shared code-path defect." }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Bare except swallowing errors in Fuyu is unrelated to the Windows ROCm torch.distributed import issue." + "reason": "Bare except in tensor conversion is unrelated to Windows ROCm import failures." }, { "left": "issue:43650", "right": "issue:46202", "accept": false, - "reason": "The noisy/unclear ADD THE DATA report is not the same as Gemma pipeline roundtrip failure." + "reason": "One is a generic placeholder issue, the other is a Gemma token roundtrip bug." }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is unrelated to the tensorflow_text backend mapping error." + "reason": "StableLm config field regression vs backend mapping error; unrelated." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "Both involve loading, but one is a legacy field in BEiT presets and the other is wrong custom-module resolution after save/load." + "reason": "Legacy preset-field loading bug vs wrong custom-module resolution after save/load." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "A Janus generate-images test failure and stale Sam3LiteText auto_mappings references are different bugs." + "reason": "Janus test failure vs Sam3LiteText auto-mapping CI breakage; different root causes." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus image-generation failure and Qwen3.5 flash-attention generation issues are unrelated." + "reason": "Janus image-generation test failure is unrelated to Qwen3.5 flash-attention generation issues." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "LoRA vocab merge collapse with tied embeddings is unrelated to Gemma4 mm_token_type_ids defaults." + "reason": "LoRA merged-vocab collapse with tied embeddings vs Gemma 4 token-type-id handling." }, { "left": "issue:43352", "right": "issue:46640", "accept": false, - "reason": "A model-specific Flash Attention 2 unsupported error is not the same as a generic FA2 issue report." + "reason": "Both mention Flash Attention, but one is a specific model support limitation and the other is a generic open issue." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Chunked-prefill cache missing and recurrent states resetting with cache are both cache-related but distinct failures." + "reason": "Cache-related symptoms, but different mechanisms: GenerationMixin cache missing vs recurrent state reset in modular_qwen3_5." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Video-classification processor selection and a missing fpn_position_embeddings attribute are different SAM/video bugs." + "reason": "Image-processor lookup in video classification is unrelated to SAM3 VisionEncoderOutput missing an attribute." }, { "left": "issue:43388", "right": "issue:43688", "accept": false, - "reason": "Gather-for-metrics label dropping and auxiliary loss normalization are separate training bugs." + "reason": "Different bugs in loss/metrics handling; no shared underlying defect." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Missing pad_token_id in Llama4Config and Qwen3-MoE sliding_window behavior are unrelated." + "reason": "Llama4 missing pad_token_id and Qwen3-MoE sliding_window behavior are separate config/model issues." }, { "left": "issue:44534", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffer junk in v5 and rotary_pct not preserved on reload are different serialization/config issues." + "reason": "General non-persistent buffer initialization bug vs GPTNeoX rotary_pct reload regression." }, { "left": "issue:35141", "right": "issue:42490", "accept": false, - "reason": "Embedding reinitialization after resize and a trailing-slash import warning are unrelated." + "reason": "Token-embedding resize/post_init weight reinit vs local-model path trailing-slash import warning." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Local dataset loading in a training script and Phi-4 multimodal model loading are not the same bug." + "reason": "Local dataset loading in a training script vs loading a specific multimodal model." }, { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "KV-cache continuation tensor indexing and Gemma4Processor missing _tokenizer are different code paths." + "reason": "KV-cache continuation indexing crash vs Gemma4Processor missing _tokenizer." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "A Python 3.13 parser/IndentationError and an s_aux=None flash_attention crash are unrelated failures." + "reason": "Python 3.13 IndentationError from a scripted function vs flash_attention s_aux=None crash." } ] }, @@ -11665,8 +11621,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -11749,196 +11705,202 @@ "issue:43854|issue:44792", "issue:39692|issue:44190" ], - "prepared_review_unit_hash": "dcc1bf1ccb3d630333028bddd485eff93f98d3aa606f7eda08ffc7c6392da45a", + "prepared_review_unit_hash": "b9c0ebeceb65d0c1a1a8b27aa79134c9748aa1c0f2095d4915a3ac2f8b08a40d", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36228, - "estimated_input_tokens": 9057, - "estimated_eval_tokens": 18370 + "serialized_chars": 36231, + "estimated_input_tokens": 9058, + "estimated_eval_tokens": 18372 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft edge looks like a true duplicate: the cluster is heterogeneous, spanning unrelated model bugs, config issues, docs regressions, and runtime crashes. If a single representative issue is needed, #47013 is the strongest anchor, but not a duplicate target.", - "confidence": 0.97, - "canonical_issue_reason": "No safe canonical duplicate target: the items do not share one underlying bug or code path.", + "summary": "No soft pair looks like a true duplicate; the overlaps are broad subsystem similarities rather than the same concrete bug. If one issue must serve as the anchor, issue:46858 is the clearest and most actionable report in the set.", + "confidence": 0.93, + "canonical_issue_reason": "issue:46858 is the most canonical-looking issue here: it has a precise failure mode, a specific trigger (static cache + torch.compile on non-CUDA backends), and an explicit root-cause description.", "canonical_pr_reason": null, - "best_issue_reason": "#47013 is the best representative issue: it is open, concrete, and has multiple inbound references with a clear failure mode.", + "best_issue_reason": "issue:46858 is the best standalone issue because it is narrowly scoped and technically specific, making it easiest to triage and fix.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45072", "right": "issue:46932", "accept": false, - "reason": "Different dtype problems in different model paths; not the same bug." + "reason": "Both mention dtype-related inference issues, but they affect different models and different code paths (bfloat16 inference vs quantized vision/audio embedders)." }, { "left": "issue:44869", "right": "issue:46832", "accept": false, - "reason": "Whisper timestamp decoding crash vs RT-DETR forward type error; unrelated." + "reason": "Whisper token-stream decoding crash and RT-DETR feature-level forward TypeError are unrelated failures in different components." }, { "left": "issue:45245", "right": "issue:46858", "accept": false, - "reason": "Category-count limit vs cache-length tensor arithmetic; different failures." + "reason": "One is a category-cardinality runtime limit, the other is a cache-length/device-tensor arithmetic bug; no shared underlying defect." }, { "left": "issue:45632", "right": "issue:46617", "accept": false, - "reason": "Cache path collision vs FileNotFoundError loading from cache; not the same root cause." + "reason": "Both involve trust_remote_code/cache behavior, but one is a leaf-directory path collision and the other is a FileNotFoundError while loading from cache." }, { "left": "issue:43673", "right": "issue:46421", "accept": false, - "reason": "Chunked-prefill cache missing vs lazy-init recompilation; related theme, different bug." + "reason": "Both mention chunked prefill/cache, but one is missing cache state in v5.0.0 while the other is spurious recompilation from lazy initialization." }, { "left": "issue:46082", "right": "issue:46697", "accept": false, - "reason": "Head-dim config validation vs image over-padding; unrelated." + "reason": "These are unrelated config vs image-processor bugs; they do not share the same failing code path." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs reference cleanup vs bad SigLIP2 outputs; not the same issue." + "reason": "Docs referring to removed pipeline APIs is not the same as SigLIP2 producing bad outputs through AutoModel/pipeline." }, { "left": "issue:29870", "right": "issue:35141", "accept": false, - "reason": "Decoder attention mask UnboundLocalError vs embedding reinit on resize; different code paths." + "reason": "Bart generate() decoder mask initialization and output-embedding reinitialization after resize are distinct model bugs." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "SAM3 video OOM vs Whisper empty-transcription crash; unrelated." + "reason": "A CUDA OOM in Sam3Video and Whisper empty-transcription generation failure are clearly different problems." }, { "left": "issue:43931", "right": "issue:44060", "accept": false, - "reason": "Weight-shape load mismatch vs incorrect tied-weights warning; different problems." + "reason": "Model weight-shape mismatch on load and an incorrect tied-weights warning are different issues in Qwen3-family models." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffer junk vs rotary_pct reload regression; not the same bug." + "reason": "Non-persistent buffer junk filling and GPTNeoX config reload losing rotary_pct are separate state/serialization bugs." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "Pipeline top_k nesting bug vs Flash Attention 2 support issue; unrelated." + "reason": "Pipeline top_k nesting behavior and XLMRoberta Flash Attention 2 support are unrelated." }, { "left": "issue:43335", "right": "issue:44855", "accept": false, - "reason": "Sparse-layer config bug vs Python 3.13 IndentationError; different failures." + "reason": "Sparse-layer construction in SwitchTransformers and a Python 3.13 IndentationError in DebertaV2Model are not the same bug." }, { "left": "issue:44242", "right": "issue:46897", "accept": false, - "reason": "Missing load-balancing loss vs Florence2 double-shift training bug; not the same code path." + "reason": "Both are training-loss related, but they concern different models and different loss-shaping mistakes." }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "Missing import error vs Qwen3.5 chat crash; unrelated." + "reason": "A missing import from image_utils and a Qwen3.5 chat crash are unrelated failures." }, { "left": "issue:44077", "right": "issue:44877", "accept": false, - "reason": "Optional post_init policy change vs strict config loading failure; different issues." + "reason": "Optional post_init handling in patchtsmixer and strict config rejection for granite_speech are different loading/configuration issues." }, { "left": "issue:45381", "right": "issue:46369", "accept": false, - "reason": "Wrong vision_position_ids vs SAM3 memory overflow; unrelated." + "reason": "A video-position-id regression and a Sam3LiteText memory overflow do not indicate the same underlying defect." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "Stale warning bug vs empty-transcription generation crash; different behavior." + "reason": "A stale warning in paged generate and a Whisper alignment/generation crash are different generate-path issues." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "torch.compile/SDPA model crash vs local dataset loading script failure; unrelated." + "reason": "Torch.compile SDPA failure in Bamba and local dataset loading failure are unrelated." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Auto device-mapping cuda assert vs category-cardinality limit; different bugs." + "reason": "An A800 cudaErrorAssert under auto device mapping is not the same as a 2^24 category limit runtime error." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "Dummy Conv2d in pixel decoder vs processor/model input-size mismatch; not the same issue." + "reason": "Sam3PixelDecoder using a dummy Conv2d and OmDet-Turbo processor/model input-size mismatch are different bugs." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Video-input regression vs KV-cache continuation crash in serve; different paths." + "reason": "A video-input regression across Qwen VL models and a serve KV-cache indexing crash are separate code paths." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "SwitchTransformers sparse-layer config and local dataset loading in the training script are unrelated." }, { "left": "issue:45588", "right": "issue:47013", "accept": false, - "reason": "Flash-attention auxiliary-state crash vs xLSTM unpacking bare tensor; unrelated." + "reason": "Flash-attention auxiliary-state handling and xLSTM return-value unpacking are different runtime errors." }, { "left": "issue:44617", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but in different models with different causes." + "reason": "Both are memory-related and both involve Sam3 variants, but they are different models with different triggers and likely different fixes." }, { "left": "issue:45381", "right": "issue:46655", "accept": false, - "reason": "Wrong video position ids vs incorrect video padding calculation; different processor bugs." + "reason": "These are both video-processor/model alignment issues, but the affected models and padding/position-id mechanics differ." }, { "left": "issue:43854", "right": "issue:44792", "accept": false, - "reason": "Model-loading unit-test failure vs Janus image-generation test failure; unrelated." + "reason": "A unit-test model-loading problem and a failed image-generation test for janus are not the same underlying bug." }, { "left": "issue:39692", "right": "issue:44190", "accept": false, - "reason": "SigLIP2 docs/example problems vs local dataset loading script issue; different areas." + "reason": "A SigLIP2 docs/model-processor mismatch and a local dataset loading failure in a training script are unrelated." } ] }, @@ -11957,8 +11919,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -12041,20 +12003,20 @@ "issue:43404|issue:45216", "issue:36246|issue:38617" ], - "prepared_review_unit_hash": "8dea7f4e9846c196f9b8249c0fe043b5069b8a86cd7a6ed73bb3c0a3567faa34", + "prepared_review_unit_hash": "5f6ca4d355cb5ba3f8c3b9c568201a154fa6e825a7d171cd49e77790466706c6", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36524, + "serialized_chars": 36521, "estimated_input_tokens": 9131, "estimated_eval_tokens": 18518 }, @@ -12063,180 +12025,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is a loose collection of unrelated Transformers issues; the soft-similarity pairs share broad themes but not the same concrete bug/change, so none should be merged as duplicates.", - "confidence": 0.88, - "canonical_issue_reason": null, + "summary": "The cluster is mostly a grab bag of unrelated Transformers issues across tokenization, generation, multimodal processing, imports, and CI/dependency problems. None of the soft-edge pairs look like true duplicates, so all are rejected. The most representative issue is the open multimodal `apply_chat_template` assistant-mask bug.", + "confidence": 0.74, + "canonical_issue_reason": "issue:44521 is the most central representative: it is open, has the most inbound references among the set, and describes a concrete multimodal chat-template bug with a clear repro surface.", "canonical_pr_reason": null, - "best_issue_reason": "No true canonical duplicate stands out; if one issue must represent the cluster, issue 44521 is the most active and well-scoped report with the strongest discussion/inbound reference signal.", + "best_issue_reason": "issue:44521 is the strongest triage candidate overall because it is active, well-discussed, and a specific user-facing bug in a core multimodal API.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "Both concern Qwen3.5/save_pretrained behavior, but they describe different failures: LoRA+extended-vocab collapse vs incorrect visual encoder keys." + "reason": "Related around Qwen3.5 saving/weights, but these are different bugs: LoRA merge with extended vocab vs incorrect visual encoder keys in `save_pretrained`." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Both mention generation, but one is about inconsistent outputs with compiled forward while the other is a stale warning in paged generate; different bugs." + "reason": "Both touch generation, but one is inconsistent compiled chunked generation outputs and the other is a stale warning about `num_return_sequences`." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Both involve resizing/token embeddings, but the regressions are different: processor/tokenizer access breakage vs Gemma4 embeddings not updating." + "reason": "Both mention model save/resize regressions, but they affect different code paths and different model families." }, { "left": "issue:44779", "right": "issue:46489", "accept": false, - "reason": "Both are DeepSeek tokenizer complaints, but they target different tokenizer regressions and likely need different fixes." + "reason": "Same tokenizer family, but one is a general DeepSeek v5 regression and the other is a DeepSeek-Coder v1-specific gap in a prior fix." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Different code paths: Qwen3OmniMoe video input validation vs transformers serve KV-cache continuation indexing." + "reason": "Different failures: Qwen3OmniMoe video-input processing vs `transformers serve` KV-cache continuation tensor indexing." }, { "left": "issue:44521", "right": "issue:45381", "accept": false, - "reason": "Both are multimodal/chat-template related, but one is assistant_masks for multimodal inputs and the other is video vision_position_ids." + "reason": "Both are multimodal-related, but one is chat-template assistant masks and the other is wrong `vision_position_ids` for Qwen2.5-VL video inputs." }, { "left": "issue:45290", "right": "issue:46752", "accept": false, - "reason": "Same API area, but not the same bug: tool-call assistant messages with no content vs empty-input IndexError." + "reason": "Both are `apply_chat_template` crashes, but the triggering inputs differ: tool-call assistant messages vs an empty list." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "CI failures from unrelated causes: stale device-override test vs auto_mappings referencing removed configs." + "reason": "Unrelated CI failures: a stale GraniteSpeech device override test vs broken `auto_mappings.py` references to removed configs." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Resize-token-embeddings bug and Whisper empty-transcription failure are unrelated code paths." + "reason": "Different bugs: embedding resize not propagating vs Whisper generation failing on empty transcription after token alignment." }, { "left": "issue:43526", "right": "issue:46697", "accept": false, - "reason": "Different image-processor bugs: label reduction returns one label vs over-padding divisible images." + "reason": "Different image-processing bugs in different models: label reduction vs over-padding in Swin2SR." }, { "left": "issue:25251", "right": "issue:29870", "accept": false, - "reason": "Pipeline top_k nesting and Bart generate decoder mask assignment are unrelated bugs." + "reason": "Distinct generation bugs: pipeline `top_k` nesting change vs Bart `decoder_attention_mask` initialization error." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and list_repo_files not forwarding revision are separate repository/client issues." + "reason": "Different APIs and failures: `AutoTokenizer` ignoring `tokenizer.json` vs `list_repo_files` not forwarding `revision`." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "XLNet CPU tensor placement bug and Fuyu broad exception handling are unrelated." + "reason": "Different root causes: XLNet device placement bug vs a bare `except:` in Fuyu tensor conversion." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Python 3.14 import failure and missing TensorFlow backend installation are different dependency problems." + "reason": "Different environment/dependency issues: Python 3.14 import failure vs TensorBoard installed without TensorFlow backend." }, { "left": "issue:43673", "right": "issue:46858", "accept": false, - "reason": "Both are cache-related, but one is missing GenerationMixin cache in chunked_prefill and the other is a device-tensor arithmetic bug in _prefill." + "reason": "Both involve cache handling, but one is missing generation cache in v5.0.0 and the other is a device-tensor slice arithmetic bug in `_prefill`." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "Both are config/loading problems, but they concern different mechanisms: trust_remote_code field loss vs unrecognized model_type." + "reason": "Different config-loading failures: missing fields with `trust_remote_code=True` vs unrecognized model type in config.json." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both involve model loading, but one is a unit-test load failure for GLM-4.7-Flash and the other is Phi-4 multimodal loading." + "reason": "Both are model-load failures, but they involve different models and different failure modes." }, { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Import slowness and unsafe sys.modules access are not the same failure." + "reason": "Different import-related problems: slow `import transformers` vs unsafe `sys.modules` access in `modeling_utils`." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "Different problems: model loading failure vs CI breakage from stale auto_mappings references." + "reason": "A model loading regression and a CI break from removed auto-mapping references are not the same bug." }, { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and output-embedding reinitialization after resize are unrelated." + "reason": "Unrelated model/runtime issues: XLMRoberta Flash Attention 2 support vs embedding reinitialization after resize when untied." }, { "left": "issue:45632", "right": "issue:46234", "accept": false, - "reason": "Both are utility/API export issues, but the missing cache path collision and missing falcon_mamba availability helper are different bugs." + "reason": "Different issues in different layers: trust-remote-code cache path collision vs missing `is_falcon_mamba_ssm_available` export." }, { "left": "issue:42796", "right": "issue:44521", "accept": false, - "reason": "Tokenizer BPE marker handling and multimodal assistant_mask generation are unrelated." + "reason": "Tokenizer BPE-marker handling and multimodal chat-template assistant masks are separate bugs." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "Stale CI device override test and model-loading regression are unrelated." + "reason": "CI test flake in GraniteSpeech is unrelated to the Phi-4-multimodal model load failure." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Package dependency/backend composition issue and generic dependency issue are too broad and not the same concrete bug." + "reason": "One is a packaging/backend install issue and the other is a vague dependency problem; not the same concrete bug." }, { "left": "issue:36246", - "right": "issue:41628", + "right": "issue:38617", "accept": false, - "reason": "Both are ImportErrors, but they expose different missing symbols from different modules." + "reason": "Both are import errors, but for different missing symbols and different modules." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer GPU broadcast NoneType error and embedding reinitialization bug are unrelated." + "reason": "Different subsystems entirely: trainer GPU broadcast of `None` vs tied/untied embedding resize reinitialization." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "Both concern Qwen3.5-style save/load behavior, but one is tied to lm_head weight tying and the other to saved checkpoint correctness." + "reason": "Different regression targets: Mistral3 weight tying vs Qwen3.5 checkpoint saving correctness." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Both are import failures, but for different missing names and unrelated modules." + "reason": "Both are import errors, but for different missing symbols and different modules." } ] }, @@ -12255,8 +12217,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -12340,20 +12302,20 @@ "issue:45335|issue:45584", "issue:44327|issue:46752" ], - "prepared_review_unit_hash": "f5381e3e152c860f845d564285f0d10f50721af7ed1ddcdb74f826c2a2f17441", + "prepared_review_unit_hash": "eda2d5cd5e4efe0877baf226d0e8ee8118d344aae65f15953a353341d19775c4", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 35951, + "serialized_chars": 35950, "estimated_input_tokens": 8988, "estimated_eval_tokens": 18232 }, @@ -12362,192 +12324,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a set of unrelated issue reports with only loose topical overlap in a few multimodal/cache/flex-attention areas. I would keep issue 44521 as the representative issue, but I do not see any soft pair that is clearly the same underlying bug or change.", - "confidence": 0.63, - "canonical_issue_reason": "Issue 44521 is the strongest representative: it is open, well-discussed, has inbound references, and describes a concrete core multimodal bug with clear reproduction and impact.", + "summary": "The cluster is heterogeneous: the soft pairs share broad Transformer/modeling themes but usually point to different subsystems, different failure modes, or different versions. None look like the same underlying bug/change, so I reject all soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "No strong canonical issue exists here; the items are not a single duplicate set, and the apparent similarities are too broad to justify merging them into one representative issue.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44521 is the best anchor for the cluster because it is the most active and actionable report, while the rest of the items are largely disjoint regressions or doc/test issues.", + "best_issue_reason": "No single issue is a good global representative because the cluster spans unrelated regressions and feature bugs across different components; any choice would be arbitrary.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "Different config-validation bugs: Llama head_dim rules vs rope key parsing from JSON lists." + "reason": "Llama head_dim validation and rope-parameter JSON list handling are unrelated config-path bugs." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Unrelated areas: Qwen3-MoE sliding window behavior vs removed Sam3LiteText auto-mappings." + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto_mappings CI breakage concern different models and code paths." }, { "left": "issue:44295", "right": "issue:46506", "accept": false, - "reason": "Different failures: position_ids buffer access vs CPU/CUDA output mismatch in PPDocLayoutV3." + "reason": "A position_ids buffer access issue and CPU/CUDA output divergence are different problems with no clear shared bug." }, { "left": "issue:41950", "right": "issue:44016", "accept": false, - "reason": "Video-classification processor lookup bug vs a notebook syntax error; not the same change." + "reason": "Video-classification processor lookup and a notebook syntax error are unrelated; one is runtime logic, the other docs/notebook content." }, { "left": "issue:43673", "right": "issue:46424", "accept": false, - "reason": "Both involve cache, but one is a missing GenerationMixin cache in chunked prefill and the other is max_cache_length handling for quantized cache only." + "reason": "Missing GenerationMixin cache in chunked_prefill is not the same as max_cache_length being honored only for quantized cache." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Separate regressions: processor/tokenizer image handling vs t5gemma resize_token_embeddings propagation." + "reason": "These both involve multimodal/composite configs, but one is a processor/tokenizer regression and the other is a resize_token_embeddings propagation bug." }, { "left": "issue:43582", "right": "issue:46932", "accept": false, - "reason": "Apple Silicon TypeError in warmup vs Gemma4 quantized vision/audio dtype casting." + "reason": "Apple Silicon TypeError in caching_allocator_warmup is unrelated to Gemma4 quantized vision/audio dtype casting." }, { "left": "issue:45083", "right": "issue:46506", "accept": false, - "reason": "Different helper bug vs cross-device numerical mismatch; no shared concrete code path." + "reason": "Qwen3 Omni MoE feature-length calculation and PPDocLayoutV3 CPU/CUDA mismatch are separate failures." }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Python version support issue vs loading a specific model failing; unrelated root causes." + "reason": "Python version incompatibility for Transformers 5.1.0 is not the same as loading a specific model failing on newer Transformers." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax fix vs audio feature extraction error; not the same bug." + "reason": "A notebook syntax error and _torch_extract_fbank_features() failure are unrelated." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Model auto-detection failure vs PIL image processor dependency issue; different subsystems." + "reason": "Unrecognized model config loading and PIL backend processors requiring torchvision are different model-loading/backend issues." }, { "left": "issue:43673", "right": "issue:44336", "accept": false, - "reason": "Generation cache bug vs ANSI escape output in loading_report; unrelated." + "reason": "chunked_prefill cache behavior and ANSI escape generation in loading_report are unrelated code paths." }, { "left": "issue:43526", "right": "issue:43901", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug vs documentation mismatch for return_all_scores." + "reason": "BeitImageProcessorFast reduce_labels bug and TextClassificationPipeline docs mismatch are not the same underlying issue." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Docs behavior mismatch vs unrecognized-model config loading; separate issues." + "reason": "Docs wording for return_all_scores and model-config recognition failure are unrelated." }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Same model family, but save/dequantization failure is distinct from FlashAttention inference failure." + "reason": "Saving dequantized Ministral/Devstral models is not the same as FlashAttention failures for Ministral-3." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both are flex_attention-related, but compiled kernel failure and AuxRequest import/version guard are separate fixes." + "reason": "Compiled flex_attention on torch>=2.9 and a wrong torch version check for AuxRequest import are distinct flex_attention issues." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "No overlap beyond being issues; unrelated requests/data text vs fbank feature extraction." + "reason": "A vague data request issue does not match a concrete audio feature extraction bug." }, { "left": "issue:43381", "right": "issue:45593", "accept": false, - "reason": "Gradient checkpointing eval-mode behavior and D-FINE auxiliary losses are unrelated." + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary losses when denoising is off are unrelated training behaviors." }, { "left": "issue:44496", "right": "issue:45759", "accept": false, - "reason": "Model recognition failure vs composite multimodal config unwrapping; different code paths." + "reason": "Unrecognized model config loading and AutoModelForCausalLM not unwrapping text_config are different composite-config problems." }, { "left": "issue:44792", "right": "issue:46424", "accept": false, - "reason": "Janus image-generation test failure vs cache-length config semantics; not the same bug." + "reason": "A failed janus image-generation test and cache max length semantics are unrelated." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Notebook syntax error vs AddedToken keyword conflict; unrelated." + "reason": "Notebook syntax error and AddedToken special-argument duplication are unrelated." }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both mention flash attention, but Ministral-3 FlashAttention failure and sink-less s_aux=None crash are distinct bugs." + "reason": "Ministral-3 FlashAttention failure and sink-less-model s_aux=None crash are both attention-related but involve different concrete failure paths." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "Python support regression vs dependency issue; too broad and not the same concrete problem." + "reason": "Python version support and a generic dependency issue do not identify the same bug." }, { "left": "issue:44521", "right": "issue:46693", "accept": false, - "reason": "Both are generation-related, but assistant_masks bug and FA2 long-context performance regression are different issues." + "reason": "all-zero assistant masks in multimodal chat templating and FA2 long-context performance regression are different generation issues." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Both are import errors, but for different missing symbols/modules and different code paths." + "reason": "Importing Qwen2_5_VLForConditionalGeneration and a nonexistent image_utils import are separate missing-export problems." }, { "left": "issue:44279", "right": "issue:46515", "accept": false, - "reason": "Dependency issue with transformers vs torch 2.4.1 incompatibility; separate root causes." + "reason": "Generic dependency trouble and torch 2.4.1 incompatibility are not the same concrete issue." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Different multimodal config bugs affecting different models and fields." + "reason": "Default vision/audio config initialization and strict-dropping of deepstack_visual_indexes are different config-field bugs." }, { "left": "issue:46123", "right": "issue:46369", "accept": false, - "reason": "Pipeline batch-finalization bug vs memory overflow in sam3-litetext; unrelated." + "reason": "Final partial batch handling in MaskGenerationPipeline is unrelated to memory overflow in sam3-litetext." }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "Resize-token-embeddings propagation and Whisper empty transcription are unrelated." + "reason": "decoder.embed_tokens resize propagation and empty-transcription Whisper generation failure are different token/decoding bugs." }, { "left": "issue:44327", "right": "issue:46752", "accept": false, - "reason": "QA top-k bounds crash vs empty-chat-template IndexError; different failure modes and code paths." + "reason": "QA top-k bounds error in decode_spans and empty-list apply_chat_template IndexError are unrelated edge cases." } ] }, @@ -12566,8 +12528,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -12651,208 +12613,208 @@ "issue:45081|issue:45907", "issue:43441|issue:45834" ], - "prepared_review_unit_hash": "51f7c456b3292e6334eac0dad2810a586373bab38156bc9839b2b2560abd5350", + "prepared_review_unit_hash": "5c4b69bf7d1f9060a4772a992d9b94edcdc506ae2fae03f55a111af56949524c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36468, - "estimated_input_tokens": 9117, - "estimated_eval_tokens": 18490 + "serialized_chars": 36480, + "estimated_input_tokens": 9120, + "estimated_eval_tokens": 18496 }, "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 issues; none of the soft pairs look like the same concrete bug/change, so I rejected them all. I used issue:43299 as the best anchor only because it is a clear regression report with a specific, reproducible loading failure.", - "confidence": 0.89, - "canonical_issue_reason": "issue:43299 is a clear regression-style bug report with concrete symptoms and version context, and is the most actionable standalone anchor in this otherwise heterogeneous set.", + "summary": "The set is highly heterogeneous: the candidate pairs mostly share only broad subsystem labels (tokenizers, configs, deepspeed, docs, multimodal models) but not the same concrete bug or change. I rejected all soft edges and found no clear canonical issue/PR representative for the whole set.", + "confidence": 0.92, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43299 has the strongest bug-report quality among the issues here: specific failure mode, explicit regression framing, and a well-defined code path.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Both involve Qwen/Qwen-like regression wording, but one is a processor/tokenizer loading issue and the other is incorrect save_pretrained key serialization; different code paths." + "reason": "Different models and failure modes: Qwen3VL MoE loading vs Qwen3.5 save_pretrained visual-encoder key serialization." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "One is a general dependency problem with transformers, the other is a PEFT minimum-version bump issue; unrelated causes and fixes." + "reason": "One is a vague dependency issue; the other is a specific PEFT version bump/release mismatch. Not the same bug." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Different model families and failures: processor.tokenizer/image-handling regression vs Deepseek tokenization output regression." + "reason": "Both are regressions, but one is processor/tokenizer API breakage and the other is Deepseek tokenization output changes. Different code paths." }, { "left": "issue:42796", "right": "issue:43650", "accept": false, - "reason": "Tokenizer BPE-marker bug versus a vague data request; not the same underlying issue." + "reason": "Completely unrelated: tokenizer BPE marker handling vs a non-informative 'ADD THE DATA' issue." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Both are docs-related, but one is a SigLIP2 example failure and the other is a TextClassificationPipeline API/docs mismatch." + "reason": "Both mention docs, but one is SigLIP2 example errors and the other is a stale TextClassificationPipeline doc note. Different fixes." }, { "left": "issue:46082", "right": "issue:46399", "accept": false, - "reason": "LlamaConfig validation around head_dim is unrelated to a Qwen3.5 kernel-registration crash." + "reason": "Different model/config bugs: LlamaConfig head_dim validation vs Qwen3.5 kernel registration failure." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "SGD arguments being ignored is a training-optimizer bug; the other is a config warning about tie_word_embeddings." + "reason": "Unrelated optimizer-argument handling vs a tie_word_embeddings warning on Qwen3.5 LoRA fine-tuning." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Blip2 dtype propagation and Qwen3.5 save_pretrained key serialization are different bugs in different models." + "reason": "Different areas: Blip2 dtype propagation on load vs Qwen3.5 save_pretrained key layout." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "A specific HybridCache import error may be one symptom, but the second issue is a broad collection of import failures; not the same concrete defect." + "reason": "Both are import-related, but one is a missing HybridCache symbol and the other is broad import failures after upgrade." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Video-classification processor selection and remote_code breakage are separate loading paths." + "reason": "Video-classification processor selection bug vs a broad remote_code regression; not the same underlying issue." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Python version compatibility and an unreleased PEFT dependency are unrelated release issues." + "reason": "Python compatibility failure vs a dependency version bump. No shared concrete bug." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and likely different tokenizer logic." + "reason": "Different tokenizer regressions affecting different model families (Deepseek vs Granite)." }, { "left": "issue:46082", "right": "issue:46512", "accept": false, - "reason": "Head-dim validation is unrelated to stale config.dtype after from_config(dtype=...)." + "reason": "Head_dim validation is unrelated to from_config(dtype=...) leaving config.dtype stale." }, { "left": "issue:44466", "right": "issue:45276", "accept": false, - "reason": "Serialized lm_head.weight depends on device versus resize_token_embeddings not propagating in Gemma4; different behaviors and fixes." + "reason": "Serialization of lm_head.weight by device is unrelated to resize_token_embeddings not updating Gemma4 embeddings." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "One is a processor/tokenizer access regression and the other is incorrect visual-encoder key saving; both Qwen3.5-related but not the same bug." + "reason": "Both touch Qwen3.5/Qwen models, but one is model loading and the other is save_pretrained visual-encoder key output. Different bugs." }, { "left": "issue:43381", "right": "issue:47000", "accept": false, - "reason": "Gradient checkpointing in eval mode and RTDetrLoss infeasible cost matrix are unrelated training/runtime problems." + "reason": "Gradient checkpointing eval-mode restriction is unrelated to RTDetrLoss infeasible cost-matrix failure." }, { "left": "issue:43650", "right": "issue:46209", "accept": false, - "reason": "A placeholder/data request is not the same as an FP8 inference failure on B200." + "reason": "Not related: placeholder issue vs FP8 inference failure on B200." }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "BigBirdTokenizer special-token registration and D-FINE auxiliary-loss logic are unrelated." + "reason": "BigBirdTokenizer special-token decode bug is unrelated to D-FINE auxiliary-loss omission." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Missing load-balancing loss when router logits are off is different from Gemma4 mm_token_type_ids defaults." + "reason": "Both concern training losses, but one is router load-balancing loss gating and the other is mm_token_type_ids defaulting in Gemma4." }, { "left": "issue:41950", "right": "issue:45823", "accept": false, - "reason": "Video pipeline processor lookup and Gemma4 PLE device placement are separate issues." + "reason": "Different multimodal model bugs: processor lookup in video pipeline vs Gemma4 PLE device placement." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Processor/tokenizer regression for custom model code is unrelated to GLM-4.7-Flash loading failures in tests." + "reason": "Different models and symptoms: Qwen3VL MoE loading vs GLM-4.7-Flash unit-test loading." }, { "left": "issue:44462", "right": "issue:45081", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and a Mistral regex-patch crash are different tokenizer-loading bugs." + "reason": "Both tokenizer-related, but AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashing are separate code paths." }, { "left": "issue:44016", "right": "issue:44291", "accept": false, - "reason": "Notebook syntax error and init_empty_weights TypeError are not the same bug." + "reason": "Notebook syntax error is unrelated to init_empty_weights/_is_hf_initialized TypeError." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "FlexAttention compilation on torch>=2.9 is unrelated to a tensorboard/TensorFlow backend packaging issue." + "reason": "Compiled flex_attention on torch>=2.9 is unrelated to a TensorBoard/TensorFlow backend import issue." }, { "left": "issue:45588", "right": "issue:47000", "accept": false, - "reason": "FlashAttention auxiliary-state crash and RTDetrLoss cost-matrix infeasibility are different model-path failures." + "reason": "Flash-attention s_aux=None crash is a different failure mode than RTDetrLoss infeasible cost matrix." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Phi4Multimodal default config initialization and Llama4 pad_token_id absence are separate config bugs." + "reason": "Both are config-related, but Phi4Multimodal default sub-config initialization and Llama4Config pad_token_id absence are distinct bugs." }, { "left": "issue:33290", "right": "issue:45656", "accept": false, - "reason": "Both mention deepspeed/optimizer behavior, but one is OOM with Adafactor and the other is double optimizer.step calls; different defects." + "reason": "DeepSpeed-related optimizer OOM and optimizer.step being called twice are not the same concrete problem." }, { "left": "issue:45081", "right": "issue:45907", "accept": false, - "reason": "Tokenizer regex crash and list_repo_files revision forwarding are unrelated repository/API bugs." + "reason": "Tokenizer regex patch crash is unrelated to list_repo_files failing to forward revision." }, { "left": "issue:43441", "right": "issue:45834", "accept": false, - "reason": "Ministral FlashAttention failure and Kosmos2.5 OCR long-input indexing are different model-specific bugs." + "reason": "Ministral FlashAttention failure and Kosmos2.5 OCR long-input index error are separate model-specific bugs." } ] }, @@ -12871,8 +12833,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -12956,20 +12918,20 @@ "issue:45083|issue:46851", "issue:44297|issue:45397" ], - "prepared_review_unit_hash": "0e5750a3167d698f63ddd20e9008f5d3269add33f1d2273def5cfa1e1e8443de", + "prepared_review_unit_hash": "f23620dd972223b62969ba0d6affdac68793eb37bcebf5d5915608558988f809", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36410, + "serialized_chars": 36409, "estimated_input_tokens": 9103, "estimated_eval_tokens": 18462 }, @@ -12978,186 +12940,186 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose embedding match across unrelated Transformers issues; I did not find any soft-edge pair that clearly describes the same underlying bug or change.", - "confidence": 0.74, - "canonical_issue_reason": "No true canonical duplicate anchor stands out because the issues span different subsystems; if one must be chosen, 46851 is the closest concrete anchor due to its reproducible position_ids regression and stronger issue signal than the more generic reports.", + "summary": "This cluster is highly heterogeneous and does not contain any clear duplicate pairs. The soft matches mostly reflect superficial title similarity around transformers v5, tokenizer, config, or loading issues, but they are different bugs or changes.", + "confidence": 0.92, + "canonical_issue_reason": "If forced to choose one issue as the representative anchor, issue 44246 is the broadest user-facing regression and has the most discussion, but the overall set is not a true duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "46851 is the strongest standalone issue: it has a clear model-specific repro, a concrete failure mode, and the most direct path to a fix among the set.", + "best_issue_reason": "Issue 44246 is the best single issue to anchor triage because it is broadly relevant, generic, and well-discussed; however, it should not be treated as a duplicate representative for the other items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44336", "right": "issue:46851", "accept": false, - "reason": "ANSI color handling in loading_report is unrelated to Qwen3.5 position_ids behavior." + "reason": "ANSI escape-code rendering in loading_report is unrelated to Qwen3.5 variable-length position_ids." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Saving quantized models and loading dequantized models on offloaded devices are different code paths and failure modes." + "reason": "Both involve quantization/offload, but one is about saving quantized models and the other about loading a dequantized model with device_map=auto." }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "processor.save_pretrained missing artifacts is unrelated to the data-filtering slowdown regression." + "reason": "processor.save_pretrained file-missing behavior is unrelated to a slowdown after upgrading Transformers." }, { "left": "issue:45859", "right": "issue:46851", "accept": false, - "reason": "Both mention Qwen3.5, but CPU-offload rotary embedding compatibility is a different bug from variable-length position_ids handling." + "reason": "Both are Qwen3.5-related, but CPU offload rotary embedding handling is a different code path from variable-length position_ids." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "auto_docstring future-annotations crash is unrelated to Gemma 4 processor loading and ReasoningEffort import issues." + "reason": "auto_docstring future-annotations crashes are unrelated to Gemma 4 processor loading/import failures." }, { "left": "issue:44261", "right": "issue:44263", "accept": false, - "reason": "A q_a_layernorm RMS epsilon precision issue is not the same as a torch.split return-value bug in GlmMoeDsaIndexer." + "reason": "These describe different model/helper bugs: rms_norm_eps precision drift vs torch.split return handling." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Missing save_pretrained files and a Zero3 from_pretrained load failure are distinct bugs." + "reason": "Missing processor.save_pretrained outputs is unrelated to Gemma-4 zero3 from_pretrained loading." }, { "left": "issue:42175", "right": "issue:43976", "accept": false, - "reason": "TensorFlow backend packaging and Python version incompatibility are unrelated setup issues." + "reason": "TensorFlow backend packaging and Python-version compatibility are different issues." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Broken auto_mappings references and leftover dummy classes are both maintenance problems, but not the same underlying defect." + "reason": "Both touch repo maintenance, but removed auto_mappings entries and leftover dummy classes are distinct problems." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Pipeline docs drift is unrelated to intermittent import-time performance." + "reason": "Documentation mismatch for return_all_scores is unrelated to intermittent import-time slowness." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "CamemBERT masked-LM output regression and Deepseek tokenizer misbehavior are separate model/tokenizer bugs." + "reason": "These are separate model regressions affecting different tokenization/model outputs." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Missing GenerationMixin cache in chunked_prefill is unrelated to CLIPVisionModel hidden_states being None." + "reason": "Chunked-prefill cache handling and CLIP hidden_states propagation are different functional bugs." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "A missing pad_token_id attribute on Llama4Config is not the same as a glm-4v-9b load failure." + "reason": "A missing pad_token_id attribute in Llama4Config is not the same as GLM-4V loading failure." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer loading in v5 and ESM2 breakage are different failures with different model surfaces." + "reason": "AutoTokenizer loading regression and ESM2 model breakage are unrelated issues." }, { "left": "issue:44016", "right": "issue:45823", "accept": false, - "reason": "Notebook syntax error is unrelated to Gemma4 device-map forward mismatch." + "reason": "Notebook syntax error and Gemma4 PLE device mismatch are unrelated." }, { "left": "issue:44188", "right": "issue:44336", "accept": false, - "reason": "torch.compile attention-kernel branching has nothing to do with ANSI codes in loading_report." + "reason": "torch.compile attention-kernel divergence has no connection to ANSI code output in loading_report." }, { "left": "issue:42796", "right": "issue:43441", "accept": false, - "reason": "Ministral tokenizer BPE handling and FlashAttention failure are both model-family issues but not the same bug." + "reason": "Tokenizer BPE marker handling and FlashAttention failure on Ministral-3 are different bugs." }, { "left": "issue:44186", "right": "issue:47000", "accept": false, - "reason": "LayoutLMv2 tokenizer padding/truncation crashes are unrelated to RTDetrLoss cost-matrix infeasibility." + "reason": "LayoutLMv2Tokenizer padding/truncation crashes are unrelated to RTDetrLoss infeasible cost matrix errors." }, { "left": "issue:43824", "right": "issue:43976", "accept": false, - "reason": "Missing Qwen2_5_VL import and Python 3.9/3.10 compatibility are unrelated." + "reason": "Missing Qwen2_5_VL export is unrelated to the Python 3.9/3.10 compatibility issue." }, { "left": "issue:44987", "right": "issue:45542", "accept": false, - "reason": "Physical-Intelligence model loading failure is unrelated to the TensorFlow/tensorboard backend mismatch." + "reason": "Model loading regression and TensorFlow backend package detection are separate problems." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Missing pad_token_id on a specific config class is not the same as trust_remote_code dropping fields from AutoConfig." + "reason": "A model config missing pad_token_id is not the same as AutoConfig losing fields under trust_remote_code." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "Terminal color output and a wrong docstring are unrelated." + "reason": "ANSI-code output in a utility and a wrong docstring are unrelated." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Missing Float8 storage type and a Qwen3.5 chat crash are not the same issue." + "reason": "Float8 storage lookup failure and Qwen3.5 chat crash are different failure modes." }, { "left": "issue:45325", "right": "issue:46851", "accept": false, - "reason": "Both involve position_ids, but one is Qwen2.5-VL rope scaling for images while the other is Qwen3.5 variable-length input handling." + "reason": "Both concern position_ids, but Qwen2.5-VL temporal scaling and Qwen3.5 variable-length inputs are different bugs." }, { "left": "issue:46655", "right": "issue:46726", "accept": false, - "reason": "SmolVLM video padding math and AutoProcessor backend detection are separate processor issues." + "reason": "Video padding calculation and missing vision backend/image processor errors are unrelated." }, { "left": "issue:46291", "right": "issue:46399", "accept": false, - "reason": "A kernels dependency/version import failure is different from a Qwen3.5 rotary-pos-emb registration crash." + "reason": "kernels v0.15 dependency versioning and missing decorator registration are different kernel-related issues." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Slow imports and an unrecognized model/config error are unrelated." + "reason": "Import latency and model config recognition failure are not the same underlying bug." }, { "left": "issue:45083", "right": "issue:46851", "accept": false, - "reason": "Qwen3_omni_moe feature-length computation is a different position_ids bug than Qwen3.5 variable-length inputs." + "reason": "These are separate Qwen-related position handling bugs in different model families and code paths." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer metadata saved incorrectly and Gemma-4 Zero3 loading failure are unrelated." + "reason": "Tokenizer save metadata mismatch is unrelated to Gemma-4 zero3 loading." } ] }, @@ -13176,8 +13138,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -13260,14 +13222,14 @@ "issue:43540|issue:46655", "issue:44496|issue:45003" ], - "prepared_review_unit_hash": "4778a3de040f2e179206f63364615df55af71d46f4cee2ad5c94fa808bd8dc38", + "prepared_review_unit_hash": "6990eb3a6df6a735b67d43d406f2b4c3cce382e9e57901247c6ae5e65dc234e8", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, @@ -13282,180 +13244,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: import/export regressions, several unrelated Qwen/MoE model-loading issues, tokenizer/pipeline bugs, and a few utility/test complaints. None of the soft pairs look like the same underlying bug or change, so I\u2019d keep them separate.", - "confidence": 0.79, - "canonical_issue_reason": "issue:42491 is the strongest anchor in the set: it is the most substantial and best-discussed item, and it matches the recurring Qwen3 MoE version-compatibility theme better than the other candidates.", + "summary": "No soft pair looks like the same underlying bug/change; they span unrelated import errors, model-loading regressions, processor bugs, and docs/runtime issues.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 45561 is the strongest standalone representative: it is open, has clear repro context, multiple inbound references, and a concrete CI/testing-utils failure mode.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42491 is the best single issue to represent the cluster only in the sense that it is the broadest and most central of the qwen-related regressions; it is not a true duplicate of the other items.", + "best_issue_reason": "Issue 45561 is the best issue to keep as a canonical reference for future duplicates because it is specific, actionable, and appears broadly relevant to test infrastructure.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41628", "right": "issue:43976", "accept": false, - "reason": "Both are import failures, but one is a missing image-processor symbol and the other is a Python-version compatibility problem; different root causes." + "reason": "Different failures: a missing AutoImageProcessor export vs a Python-version compatibility issue; not the same code-path bug." }, { "left": "issue:43784", "right": "issue:45542", "accept": false, - "reason": "Both involve import-time backend issues, but one is a missing nn symbol in sentence-transformers and the other is a TensorFlow/tensorboard backend error." + "reason": "Both are import-time errors, but they concern different missing backends/symbols in different integrations ('nn' vs undefined tf backend)." }, { "left": "issue:46399", "right": "issue:46851", "accept": false, - "reason": "Same model family, but different code paths: kernel registration/decorator failure versus variable-length position_ids handling." + "reason": "Both involve Qwen3.5, but one is kernel registration/use_kernels and the other is variable-length position_ids; distinct bugs." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "Completely different areas: xdist file-race in testing utils versus a stale generate() warning." + "reason": "Test utility race on captured_info.txt is unrelated to a stale generation warning for num_return_sequences." }, { "left": "issue:43540", "right": "issue:46752", "accept": false, - "reason": "Different features and failures: Qwen3OmniMoe video input processing versus empty chat-template input handling." + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to apply_chat_template crashing on an empty list." }, { "left": "issue:43526", "right": "issue:45003", "accept": false, - "reason": "Unrelated bugs in different subsystems: BEiT image label reduction versus sys.modules access in modeling_utils." + "reason": "BeitImageProcessorFast label reduction bug and unsafe sys.modules access in modeling_utils are different subsystems and failures." }, { "left": "issue:45103", "right": "issue:46399", "accept": false, - "reason": "Docstring/annotation parsing crash versus Qwen3.5 kernel registration failure; no shared code path." + "reason": "Auto-docstring annotation handling is unrelated to Qwen3.5 rotary kernel registration." }, { "left": "issue:45072", "right": "issue:46829", "accept": false, - "reason": "Different model families and symptoms: dtype mismatch in inference versus wrong tensor returned in output dataclass." + "reason": "Different model families and symptoms: dtype mismatch in bfloat16 inference vs hidden_states output wiring." }, { "left": "issue:46614", "right": "issue:46665", "accept": false, - "reason": "Tokenizer backend caching bug versus inconsistent from_pretrained behavior inside Python classes; not the same defect." + "reason": "A stale phonemizer backend cache bug is not the same as inconsistent from_pretrained behavior inside Python classes." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "Position_ids buffer access bug is unrelated to gemma-4 ZeRO3 from_pretrained behavior." + "reason": "position_ids buffer reading error and gemma-4 ZeRO3 from_pretrained failure are distinct." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Both are Qwen-related, but one is a sliding_window bug in Qwen3-MoE and the other is a chat-loading crash in Qwen3.5." + "reason": "Qwen3-MoE sliding_window handling and Qwen3.5 chat crash are different code paths." }, { "left": "issue:45588", "right": "issue:46851", "accept": false, - "reason": "Flash-attention sink handling and variable-length position_ids are different failures in different paths." + "reason": "Flash-attention s_aux=None crash is unrelated to variable-length position_ids handling." }, { - "left": "issue:44242", + "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Load-balancing loss computation and Gemma eos_token_id config mismatch are unrelated." + "reason": "Import latency and eos_token_id config mismatch are unrelated issues." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both mention Qwen3 MoE/version changes, but one is a LoRA compatibility problem and the other is a Qwen3VL MoE loading regression; not the same concrete bug." + "reason": "Both mention Qwen3 MoE loading, but one is a LoRA/main-branch compatibility problem and the other is a Qwen3VL-MoE regression; too different to merge." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "ANSI output formatting bug versus GLM-5 RoPE implementation discussion; no shared underlying issue." + "reason": "ANSI-code cleanup in loading_report is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:45003", "right": "issue:46821", "accept": false, - "reason": "Sys.modules access bug and TimesFM window_size AttributeError are unrelated." + "reason": "Unsafe sys.modules access and TimesFM window_size AttributeError are unrelated." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Both mention missing auxiliary loss behavior, but they affect different models and different mechanisms." + "reason": "Both concern missing auxiliary losses, but for different models and different conditions; not the same concrete bug." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline cache loading and AttentionInterface.register behavior are separate concerns." + "reason": "Offline tokenizer cache loading and AttentionInterface.register behavior change are unrelated." }, { "left": "issue:43550", "right": "issue:46506", "accept": false, - "reason": "Different models, platforms, and failure modes: torch.compile/SDPA crash versus CPU/CUDA output mismatch." + "reason": "torch.compile/SDPA failure in Bamba is unrelated to CPU-vs-CUDA output differences in PPDocLayoutV3." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "Tokenizer correctness regression versus embedding resize propagation bug; different code paths." + "reason": "Deepseek tokenizer regression and resize_token_embeddings not updating output_embeddings are different bugs." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but for different packages and symbols, so they are not the same bug." + "reason": "Missing Qwen2.5 VL image processor export and missing is_offline_mode export are separate import errors." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Missing Qwen2_5_VLImageProcessor export and missing HybridCache export are separate import-name regressions." + "reason": "Both are missing-name import errors, but they involve different exported symbols and modules." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "Documentation mismatch for a pipeline option versus a low-level sys.modules access bug." + "reason": "Docs mismatch for return_all_scores and sys.modules access are unrelated." }, { "left": "issue:44451", "right": "issue:46752", "accept": false, - "reason": "Model loading failure for ScandiBERT is unrelated to empty chat-template handling." + "reason": "Model-loading failure for ScandiBERT is unrelated to apply_chat_template on an empty list." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MoE weight conversion/loading and a DeepSpeed ZeRO3 deque error are different loader failures." + "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed and a ZeRO3 deque IndexError are different DeepSpeed failures." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification pipeline misrouting to image processors is unrelated to fbank feature extraction." + "reason": "Video-classification looking for image processors and a torch_extract_fbank_features issue are different pipeline bugs." }, { "left": "issue:43540", "right": "issue:46655", "accept": false, - "reason": "Both involve video processors, but they are different models and different bugs: input-processing ValueError versus bad padding calculation." + "reason": "Both involve video processing, but they affect different models and different failure mechanisms (ValueError vs padding calculation)." }, { "left": "issue:44496", "right": "issue:45003", "accept": false, - "reason": "Unrecognized model config and unsafe sys.modules access are not the same issue." + "reason": "Unrecognized model_type in config loading is unrelated to unsafe sys.modules access." } ] }, @@ -13474,8 +13436,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -13556,190 +13518,190 @@ "issue:44368|issue:44977", "issue:43653|issue:44162" ], - "prepared_review_unit_hash": "033e9db7f60bd7cf3f70e9d6883adb9502fc384aa0531f31767521c576000777", + "prepared_review_unit_hash": "6313c86199f85de96c9986887615b866d761b883760265b4df737cd119c4456a", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 35789, - "estimated_input_tokens": 8948, - "estimated_eval_tokens": 18152 + "serialized_chars": 35794, + "estimated_input_tokens": 8949, + "estimated_eval_tokens": 18154 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: it mixes unrelated bugs across models, processors, training, logging, and performance. None of the soft pairs looks like the same underlying defect, so no merges are justified.", - "confidence": 0.93, - "canonical_issue_reason": "No true canonical issue stands out because the items are not duplicates of one another; if forced to anchor the cluster, issue 45561 is the most actionable report due to its clear reproduction and discussion history.", + "summary": "The proposed soft links are all false positives: the issues span unrelated models, processors, training bugs, and infrastructure problems with only superficial wording overlap.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the candidates are heterogeneous and the soft links are weak or superficial.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45561 is the best single anchor only because it is a well-scoped, reproducible bug with inbound references and active discussion, but it does not represent the rest of the mixed set.", + "best_issue_reason": "issue:45561 is the most active and well-specified standalone bug in the set, but it does not serve as a duplicate anchor for the rest.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Both touch Qwen 3.5 multimodal config handling, but one is a missing config field under strict deserialization while the other is AutoModelForCausalLM.from_config not unwrapping text_config; different code paths." + "reason": "Both involve Qwen multimodal config/model loading, but they hit different bugs: strict-config field dropping vs unwrapping text_config in from_config." }, { "left": "issue:43262", "right": "issue:44246", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting is unrelated to import-time slowness in transformers." + "reason": "Audio chat-template sampling-rate handling and import-time slowdown are unrelated issues." }, { "left": "issue:44479", "right": "issue:46655", "accept": false, - "reason": "Both are video-related, but they affect different models and failures: Qwen video regression vs SmolVLM padding calculation." + "reason": "Both are video regressions, but for different models and different failure causes (general Qwen video regression vs SmolVLM padding calculation)." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "Unsafe sys.modules access and an xdist race on captured_info.txt are separate bugs in different subsystems." + "reason": "sys.modules access safety and a pytest-xdist file race are distinct subsystems and bugs." }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 breakage and Gemma 4 mm_token_type_ids defaulting are unrelated model-specific issues." + "reason": "ESM2 breakage and Gemma 4 mm_token_type_ids defaults are unrelated model-specific issues." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "Registering position_ids as a buffer causing read errors is not the same as a slowdown after upgrading transformers." + "reason": "A position_ids buffer access error and a post-upgrade performance regression are not the same bug." }, { "left": "issue:45859", "right": "issue:46399", "accept": false, - "reason": "Both involve Qwen3.5, but CPU-offload rotary embedding incompatibility and a missing decorator for kernel registration are different concrete bugs." + "reason": "Both are Qwen3.5-related, but CPU offload rotary embedding support and apply_rotary_pos_emb/kernel registration are different code paths." }, { "left": "issue:45292", "right": "issue:47000", "accept": false, - "reason": "resize_token_embeddings/output_embeddings mismatch is unrelated to RTDetrLoss crashing from an infeasible cost matrix." + "reason": "Embedding-resize propagation and RTDetr loss infeasible-cost-matrix crashes are unrelated." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow and stale auto_mappings references are unrelated failures." + "reason": "DeepSpeed ZeRO3 deque underflow and stale auto_mappings config references are different problems." }, { "left": "issue:43475", "right": "issue:46369", "accept": false, - "reason": "A missing Sam3VisionEncoderOutput attribute is not the same as memory overflow in sam3-litetext." + "reason": "Missing fpn_position_embeddings and memory overflow in sam3-litetext are different SAM3 issues." }, { "left": "issue:43295", "right": "issue:45292", "accept": false, - "reason": "Custom processor.tokenizer/image handling regression is unrelated to resize_token_embeddings not updating output embeddings." + "reason": "Custom processor/tokenizer regression and output embedding resize propagation are different APIs and failure modes." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "A missing Qwen2_5_VLImageProcessor import and a DeepSpeed/BertModel index error are different issues." + "reason": "Qwen2.5-VL import failure and BertModel ZeRO3 initialization IndexError are unrelated." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Tied-weight serialization depending on device is not the same as t5gemma resize_token_embeddings failing to update decoder.embed_tokens." + "reason": "Both mention embeddings/tied weights, but one is serialization inconsistency and the other is resize_token_embeddings not updating decoder embeddings." }, { "left": "issue:44623", "right": "issue:46726", "accept": false, - "reason": "Missing files in processor.save_pretrained and AutoProcessor's backend-missing error are different save/load problems." + "reason": "Missing files on save_pretrained and missing vision backend handling in AutoProcessor are different bugs." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Both mention strict config, but one blocks granite_speech loading while the other silently drops a missing Qwen3_5MoeVisionConfig field; not the same bug." + "reason": "Strict config loading for granite_speech and a dropped Qwen config field are separate config-handling issues." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Missing load-balancing loss when router logits are off is unrelated to LWDetrImageLoss crashing under float16 AMP on CUDA." + "reason": "Load-balancing loss omission and an AMP/CUDA crash in LwDetrImageLoss are unrelated." }, { "left": "issue:43526", "right": "issue:46897", "accept": false, - "reason": "BEiT reduce_labels returning only one label is unrelated to Florence2's training-loss double-shift bug." + "reason": "Beit label reduction and Florence2 training-loss double-shift are different preprocessing/training bugs." }, { "left": "issue:44038", "right": "issue:46506", "accept": false, - "reason": "A Qwen3-VL-Moe/transformers 5 issue is not the same as CPU-vs-CUDA differences in PPDocLayoutV3." + "reason": "A transformers/Qwen3-VL-Moe bug and a CPU/CUDA output discrepancy in PPDocLayoutV3 are not the same issue." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Model loading failure with transformers>=5.1.0 and protobuf error masking are different bugs." + "reason": "Model loading failure on transformers>=5.1.0 and tokenizer error masking when protobuf is missing are different failure modes." }, { "left": "issue:29870", "right": "issue:30333", "accept": false, - "reason": "Bart generate() unboundlocalerror and MLFlow job-status reporting are unrelated." + "reason": "Bart generate() UnboundLocalError and MLflow job-status reporting are unrelated." }, { "left": "issue:42199", "right": "issue:46851", "accept": false, - "reason": "DETR cardinality/background-class handling and Qwen3.5 variable-length position_ids are different code paths." + "reason": "DETR cardinality logic and Qwen3.5 variable-length position_ids are different model bugs." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "Unsupported Flash Attention 2.0 for NemotronH is not the same as the generic Flash Attention 2.0 item." + "reason": "Unsupported Flash Attention 2.0 for Nemotron and a generic Flash Attention topic are not the same concrete bug." }, { "left": "issue:43526", "right": "issue:46894", "accept": false, - "reason": "BEiT label reduction and CohereASR training-loss double-shift are unrelated despite both being training/preprocessing bugs." + "reason": "Beit label reduction and CohereASR training-loss double-shift are distinct issues." }, { "left": "issue:44060", "right": "issue:44077", "accept": false, - "reason": "An incorrect tied-weights warning and a patchtsmixer post_init policy change are unrelated." + "reason": "Incorrect tied-weights warning and patchtsmixer post_init policy are unrelated." }, { "left": "issue:44368", "right": "issue:44977", "accept": false, - "reason": "Qwen3.5 tied-embeddings warnings and Qwen3.5 flash-attention generation failure are different concrete issues." + "reason": "A tie_word_embeddings warning in LoRA fine-tuning and a flash-attention generation regression are different problems." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "BigBirdTokenizer mask-token registration and ESM2 breakage are unrelated." + "reason": "BigBird tokenizer special-token decoding and ESM2 breakage are unrelated." } ] }, @@ -13845,220 +13807,220 @@ "issue:44623|issue:45656", "issue:44871|issue:46132" ], - "prepared_review_unit_hash": "7539dcd3f3f1a188afe6bd15332580a6bb5c633d562c47cd922aa327b2999505", + "prepared_review_unit_hash": "04d2cf586e4c8f81e12c57d24455d22451a46c8fa0abed39abd8cf05ef27fd70", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37152, - "estimated_input_tokens": 9288, - "estimated_eval_tokens": 18832 + "serialized_chars": 37159, + "estimated_input_tokens": 9290, + "estimated_eval_tokens": 18836 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The proposed soft pairs are mostly broad thematic overlaps rather than the same concrete bug or change. I\u2019d reject all of them as duplicates; the set spans unrelated tokenizer, model-loading, cache, attention, and training issues.", - "confidence": 0.92, - "canonical_issue_reason": null, + "summary": "The set is mostly heterogeneous, with one small tokenizer/processor serialization subgroup and many unrelated issues (vision, cache, deepspeed, Gemma, RT-DETR, etc.). No soft edges look like the same underlying bug/change, so all candidate merges are rejected. Issue #44297 is the best representative of the tokenizer serialization subgroup.", + "confidence": 0.34, + "canonical_issue_reason": "#44297 is the clearest core bug in the tokenizer save/load path: it directly describes incorrect tokenizer metadata being written, which can plausibly underlie nearby tokenizer serialization/loading complaints.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the strongest standalone bug report here: it states a clear regression, names multiple affected model families, and points to a concrete video-input failure path.", + "best_issue_reason": "#44297 best captures the most coherent subcluster around tokenizer/processor persistence and loading; it is concrete, actionable, and closest to the shared failure mode among the few related items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Video-input regression in Qwen/VL models and a Windows ROCm import/dependency issue are unrelated code paths." + "reason": "Different subsystems and failure modes: video-input regression in Qwen VL models vs Windows/ROCm import/type errors." }, { "left": "issue:45412", "right": "issue:46123", "accept": false, - "reason": "RT-DETR memory not being released and MaskGenerationPipeline dropping a final partial batch are different bugs in different components." + "reason": "Memory reclamation in RT-DETR is unrelated to MaskGenerationPipeline's partial-batch `is_last` logic." }, { "left": "issue:45397", "right": "issue:45656", "accept": false, - "reason": "A zero3 from_pretrained failure and a double optimizer-step problem under deepspeed are not the same underlying issue." + "reason": "Gemma-4 Zero3 loading and deepspeed optimizer double-step are different bugs in different code paths." }, { "left": "issue:42886", "right": "issue:45229", "accept": false, - "reason": "Offline tokenizer cache loading and Gemma4 multi-GPU CUDA OOM are unrelated symptoms and subsystems." + "reason": "Offline tokenizer cache loading is unrelated to Gemma4 multi-GPU CUDA OOM." }, { "left": "issue:43065", "right": "issue:43408", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d in the pixel decoder and the other is a model-type/config mismatch warning." + "reason": "One is a dummy Conv2d implementation detail; the other is a model-type warning during loading." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "These concern different models and failures: a missing vision-encoder attribute versus a Qwen3OmniMoe video-processing ValueError." + "reason": "SAM3 vision encoder attribute error and Qwen3OmniMoe video-processing ValueError are separate model-specific failures." }, { "left": "issue:44877", "right": "issue:46170", "accept": false, - "reason": "Strict config loading for granite_speech and a Windows ROCm/torch.distributed import issue do not share the same bug." + "reason": "Strict config loading for granite_speech is unrelated to the Windows ROCm distributed import issue." }, { "left": "issue:43723", "right": "issue:46710", "accept": false, - "reason": "Tokenizer loading in v5 and a DeepSeek accuracy regression after upgrade are distinct problems." + "reason": "Tokenizer loading failure in v5 and an accuracy regression after upgrade are not the same concrete bug." }, { "left": "issue:46170", "right": "issue:46832", "accept": false, - "reason": "Import failure on Windows/ROCm and an RTDetr forward shape/feature-level error are different code paths." + "reason": "Platform import/dependency failure and RT-DETR feature-level mismatch are unrelated." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and a tie_word_embeddings warning in ms-swift fine-tuning are unrelated." + "reason": "Qwen3-VL weight shape mismatch and a tie_word_embeddings warning in Qwen3.5 are different issues." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "A torch.compile attention-kernel divergence bug is not the same as a Gemma pipeline roundtrip failure on the \u2581 token." + "reason": "torch.compile attention-kernel branching divergence is unrelated to Gemma pipeline roundtrip tokenization failure." }, { "left": "issue:44368", "right": "issue:45563", "accept": false, - "reason": "One is a model config warning, the other is a stale num_return_sequences warning in paged generate; not the same bug." + "reason": "A configuration warning about tied embeddings does not match a stale generate warning for num_return_sequences." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "GlmMoeDsa stale indexer cache on second forward and Gemma4 mm_token_type_ids defaulting are different defects." + "reason": "GlmMoeDsa cache staleness on second forward is unrelated to Gemma4 mm_token_type_ids defaults." }, { "left": "issue:44483", "right": "issue:46202", "accept": false, - "reason": "A /v1/chat/completions request rejection and a pipeline tokenizer roundtrip issue are not the same underlying problem." + "reason": "Chat/completions API request rejection is unrelated to Gemma pipeline text roundtrip behavior." }, { "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "An isolated reproduction of another issue and a Gemma roundtrip failure are unrelated." + "reason": "An isolated reproduction of another issue and a Gemma roundtrip bug are not the same underlying defect." }, { "left": "issue:44188", "right": "issue:46421", "accept": false, - "reason": "Both mention cache/compile, but one is attention-kernel divergence and the other is lazy cache initialization causing recompilation." + "reason": "Attention kernel branching divergence is unrelated to cache lazy initialization causing recompilation." }, { "left": "issue:45987", "right": "issue:46082", "accept": false, - "reason": "StaticCache returning a shape-(1,) tensor and LlamaConfig validating head_dim divisibility are separate contract/validation issues." + "reason": "StaticCache return-type bug and LlamaConfig head_dim validation are different code paths with different fixes." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "A stale cache bug on second forward and a flash_attention s_aux=None crash are different failures." + "reason": "GlmMoeDsa stale indexer cache and flash_attention s_aux=None AttributeError are unrelated failures." }, { "left": "issue:45593", "right": "issue:47068", "accept": false, - "reason": "Missing auxiliary losses in D-FINE and a CUDA-tensor conditional causing host sync in deepspeed loss code are unrelated." + "reason": "D-FINE auxiliary-loss behavior and a deepspeed host-synchronization concern are not the same bug." }, { "left": "issue:45685", "right": "issue:46431", "accept": false, - "reason": "An MPS histogram integer limitation and a missing chat template error are different problems." + "reason": "MPS histogram support in MoE is unrelated to apply_chat_template missing-template errors." }, { "left": "issue:45103", "right": "issue:46858", "accept": false, - "reason": "auto_docstring crashes with future annotations, while _prefill has a static-cache tensor/device arithmetic bug." + "reason": "Auto-docstring annotation handling and cache.get_seq_length device-tensor arithmetic are separate issues." }, { "left": "issue:44361", "right": "issue:44749", "accept": false, - "reason": "MLukeTokenizer AttributeError and a Transformers upgrade performance regression are not the same issue." + "reason": "MLukeTokenizer AttributeError on tasks and a performance regression after upgrade do not share a concrete root cause." }, { "left": "issue:43065", "right": "issue:46123", "accept": false, - "reason": "SAM3 pixel-decoder scaffolding and MaskGenerationPipeline partial-batch handling are unrelated." + "reason": "SAM3 dummy Conv2d and MaskGenerationPipeline partial-batch dropping results are unrelated." }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch and processor save_pretrained missing files are both serialization-related, but not the same concrete bug." + "reason": "These are the closest pair, but one is tokenizer metadata mismatch while the other is processor save_pretrained missing files; related area, not the same bug." }, { "left": "issue:43994", "right": "issue:46726", "accept": false, - "reason": "Nonsensical outputs from SigLIP2 and an AutoProcessor missing-backend error are different failure modes." + "reason": "SigLIP2 nonsensical outputs and AutoProcessor missing-vision-backend handling are distinct model-loading/processing issues." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "A slowdown after upgrading Transformers and a deepspeed optimizer-step duplication bug are unrelated." + "reason": "A broad slowdown regression after upgrade is not the same as deepspeed calling optimizer.step twice." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "resume_from_checkpoint key mismatch and post-upgrade filtering slowdown are different issues." + "reason": "resume_from_checkpoint key mismatch and a training slowdown regression are unrelated." }, { "left": "issue:46752", "right": "issue:47020", "accept": false, - "reason": "Empty chat-template input IndexError and XLMRobertaTokenizer passing a dict to Unigram are unrelated tokenizer bugs." + "reason": "Empty-list chat-template crash and XLMRobertaTokenizer/Unigram construction error are different tokenizer bugs." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "Custom model initialization breaking in v5 and ESM2 being broken are separate model-specific regressions." + "reason": "Custom model initialization breakage in Transformers 5 and ESM2 being broken are too broad and not the same defect." }, { "left": "issue:44623", "right": "issue:45656", "accept": false, - "reason": "Processor save_pretrained file emission and double optimizer stepping under deepspeed are unrelated." + "reason": "Processor save_pretrained missing files has nothing to do with deepspeed optimizer stepping twice." }, { "left": "issue:44871", "right": "issue:46132", "accept": false, - "reason": "Gemma eos_token_id inconsistency and AttentionInterface.register changing registered function behavior are different concerns." + "reason": "Gemma eos_token_id inconsistency and AttentionInterface.register behavior changes are separate API/config issues." } ] }, @@ -14077,8 +14039,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -14148,8 +14110,8 @@ "issue:44038|issue:44623", "issue:46877|issue:47000", "issue:41628|issue:44933", - "issue:46463|issue:46877", "issue:42994|issue:43525", + "issue:46463|issue:46877", "issue:44869|issue:45356", "issue:43723|issue:46326", "issue:42371|issue:45092", @@ -14164,226 +14126,226 @@ "issue:44661|issue:47059", "issue:44038|issue:45310" ], - "prepared_review_unit_hash": "e2fde61fdb216a42461795051a1231b707c1e2a03a08fdb6d733c9b088b1b86a", + "prepared_review_unit_hash": "bc5b47989b019a6833fe3bd4a5790fb30c1a5ee1a1727c8abe3a038bb7f2a2f0", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36645, - "estimated_input_tokens": 9162, - "estimated_eval_tokens": 18580 + "serialized_chars": 36639, + "estimated_input_tokens": 9160, + "estimated_eval_tokens": 18576 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of Transformers issue reports centered on API regressions, model-loading failures, and missing/changed attributes. A few Qwen3.5 MoE reports and AutoProcessor backend-missing reports look duplicative, but most pairs are only superficially similar and should stay separate.", - "confidence": 0.43, - "canonical_issue_reason": "issue:45310 is the clearest representative of the cluster\u2019s strongest duplicate-like thread: a Qwen3.5 MoE from_pretrained regression in Transformers>=5.4.0, with closely matching follow-ups around the same model/config failure.", + "summary": "This cluster is mostly a loose collection of unrelated Transformers bugs, with one clear duplicate family around Qwen3.5-Moe / Qwen3.5-MoeVision from_pretrained failures. The rest span different models, APIs, and subsystems, so they should not be merged together.", + "confidence": 0.77, + "canonical_issue_reason": "Issue 45310 is the best representative because it is the clearest, most general description of the recurring Qwen3.5-Moe loading regression (`transformers>=5.4.0` + `from_pretrained` failure), which matches the closest duplicate set in this cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 has the best mix of specificity and duplicate linkage; it captures one of the few concrete recurring bugs in the set rather than a vague API/import complaint.", + "best_issue_reason": "Issue 45310 is the strongest global representative: it names the concrete model family, the affected version range, and the failing load path, making it the most useful anchor for the duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45375", "right": "issue:45478", "accept": true, - "reason": "Same Qwen3.5 MoE from_pretrained/config regression; titles and symptoms line up closely enough to be one bug." + "reason": "Both reports point to the same Qwen3.5-Moe/Vision `from_pretrained` regression, with the missing `deepstack_visual_indexes` field being a concrete manifestation of the loading failure." }, { "left": "issue:42994", "right": "issue:43482", "accept": false, - "reason": "Both involve quantization/serialization, but one is saving failure and the other is GGUF loading failure; different code paths." + "reason": "These are different bugs: one is quantized model saving, the other is GGUF loading under Transformers v5." }, { "left": "issue:44964", "right": "issue:45137", "accept": false, - "reason": "Completely different failures: multimodal model loading vs DeepSpeed ZeRO3 deque underflow." + "reason": "Different failure modes and subsystems: multimodal model loading versus DeepSpeed ZeRO3 deque handling." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "Both mention missing fields/attributes, but they involve different models and different mechanisms (SAM output vs AutoConfig trust_remote_code)." + "reason": "One is a SAM 3 Vision encoder attribute error; the other is a generic `trust_remote_code` config field-loss issue. Similar symptom class, but not the same bug." }, { "left": "issue:44188", "right": "issue:46851", "accept": false, - "reason": "Different bugs: torch.compile attention-kernel branching vs Qwen3.5 variable-length position_ids handling." + "reason": "Both involve generation/attention behavior, but they are distinct code paths and symptoms: compile-time kernel branching versus variable-length position IDs." }, { "left": "issue:43475", "right": "issue:46821", "accept": false, - "reason": "Both are attribute errors, but on unrelated model internals and unrelated entry points." + "reason": "Different models and different missing attributes (`fpn_position_embeddings` vs `window_size`)." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Both touch generation internals, but one is sync_gpus kwargs update and the other is chunked_prefill cache exposure; not the same bug." + "reason": "Both touch generation internals, but one is an `_update_model_kwargs_for_generation` ordering bug and the other is a missing `GenerationMixin` cache issue in chunked prefill." }, { "left": "issue:39692", "right": "issue:46726", "accept": false, - "reason": "Doc example/model mismatch and quantization failure versus missing-vision-backend processor error; different underlying issues." + "reason": "The former is a SigLIP2 documentation/example problem; the latter is a runtime AutoProcessor backend-missing error." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "PyTorch-missing import error is unrelated to a loading hang in Sentence Transformers." + "reason": "Missing PyTorch in the environment is not the same as Sentence Transformers getting stuck while loading." }, { "left": "issue:43986", "right": "issue:46726", - "accept": true, - "reason": "Both report confusing AutoProcessor failures when the required vision backend is missing; same underlying missing-backend handling issue." + "accept": false, + "reason": "Both involve processor/backend dependencies, but the concrete issues differ: video-model AutoProcessor crash without torchvision versus missing-vision-backend image processor handling." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "No concrete overlap: one is an empty bug report, the other is a ZeRO3/from_pretrained issue for Gemma-4." + "reason": "Too generic to merge: one is an empty bug report, the other is a specific Gemma-4 ZeRO3 from_pretrained failure." }, { "left": "issue:45310", "right": "issue:45375", "accept": true, - "reason": "Same Qwen3.5 MoE Transformers v5 regression; 45375 spells out the missing strict-field root cause behind the from_pretrained error." + "reason": "These are the same underlying Qwen3.5-Moe loading/config regression; 45375 provides a specific missing-field symptom for the broader from_pretrained failure described in 45310." }, { "left": "issue:44373", "right": "issue:46851", "accept": false, - "reason": "Docstring typo for position_ids is unrelated to a runtime variable-length input bug." + "reason": "A docstring correction is unrelated to a variable-length input runtime bug." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Different areas: Qwen3-VL-Moe v5 bug versus processor.save_pretrained missing files." + "reason": "Qwen3-VL-Moe loading failure and `processor.save_pretrained(...)` missing files are different problems." }, { "left": "issue:46877", "right": "issue:47000", "accept": false, - "reason": "Doc-lint stdout noise and RTDetrLoss infeasible cost matrix are unrelated." + "reason": "Import-time doc-lint output and RTDetrLoss infeasibility are unrelated." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both are import-related, but they are different missing symbols from different modules and likely different fixes." + "reason": "Both are import-related, but they concern different missing symbols in different modules (`AutoImageProcessor` vs image_utils export)." }, { - "left": "issue:46463", - "right": "issue:46877", + "left": "issue:42994", + "right": "issue:43525", "accept": false, - "reason": "Incorrect T5 validation error message is unrelated to auto_docstring stdout diagnostics." + "reason": "Quantized saving failure and missing `pad_token_id` on Llama4Config are unrelated." }, { - "left": "issue:42994", - "right": "issue:43525", + "left": "issue:46463", + "right": "issue:46877", "accept": false, - "reason": "Quantized model saving failure and missing pad_token_id on Llama4Config are different bugs." + "reason": "Incorrect decoder-input validation messaging is unrelated to import-time docstring diagnostics." }, { "left": "issue:44869", "right": "issue:45356", "accept": false, - "reason": "Whisper timestamp decoding crash and Kimi tokenizer regression are separate tokenizer bugs." + "reason": "Whisper timestamp decoding crash and Kimi tokenizer codec regression are different tokenizer bugs." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "Tokenizer loading regression and mistral_common continue_final_message bug are different code paths." + "reason": "Tokenizer loading in v5 and `continue_final_message` behavior for mistral_common are separate issues." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 settings deprecation and InternVL2 meta-init incompatibility are unrelated." + "reason": "TF32 API settings and InternVL2 meta-initialization incompatibility are unrelated." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization offloading behavior and pipeline-parallel output correctness are different problems." + "reason": "Quantization offloading behavior and pipeline-parallelism output corruption are different runtime issues." }, { "left": "issue:45561", "right": "issue:47059", "accept": false, - "reason": "pytest-xdist race in testing utils is unrelated to a CLI incompatibility with huggingface_hub." + "reason": "A pytest-xdist race in testing utils is unrelated to a CLI completion attribute error from huggingface_hub." }, { "left": "issue:43784", "right": "issue:46399", "accept": false, - "reason": "Sentence-transformers import failure and Qwen3.5 kernel-registration crash are different failures." + "reason": "Missing `nn` import in sentence-transformers and a Qwen3.5 kernel-registration decorator error are different bugs." }, { "left": "issue:41628", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different public symbols and likely different API breaks." + "reason": "Different missing imports: `AutoImageProcessor` versus `HybridCache`." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallel output mismatch is unrelated to a Qwen3.5 image-specific infinite loop." + "reason": "Pipeline parallelism output issues and Qwen3.5 infinite-loop behavior are not the same failure." }, { "left": "issue:45406", "right": "issue:46202", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and Gemma pipeline roundtrip on \u2581 are different issues." + "reason": "Gemma4 processor missing `_tokenizer` and Gemma pipeline roundtrip failure on a special token are unrelated." }, { "left": "issue:45137", "right": "issue:47000", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error and RTDetrLoss infeasible cost matrix are unrelated runtime failures." + "reason": "DeepSpeed ZeRO3 deque underflow and RTDetrLoss cost-matrix infeasibility are different error sources." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "GenerationMixin import breakage and AutoImageProcessor import breakage are separate missing-export issues." + "reason": "Different missing imports from different APIs (`GenerationMixin` vs `AutoImageProcessor`)." }, { "left": "issue:44661", "right": "issue:47059", "accept": false, - "reason": "Model-registration logic and CLI completion compatibility are unrelated." + "reason": "A model-adder failure in tokenizer mapping is unrelated to CLI completion breakage." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Both concern Qwen models on v5, but they are different model families and different failure modes." + "reason": "Both are Qwen-family loading bugs, but the model/version and symptom details are different enough that they do not clearly describe the same underlying defect." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet, and the accepted soft-edge verdicts are conservative enough. The strongest duplicate thread around Qwen3.5 MoE v5 regressions is well supported, and the AutoProcessor/backend pair is a plausible same-bug linkage rather than an overreach." + "feedback": "The summary is grounded in the packet: the Qwen3.5-Moe/Qwen3.5-MoeVision from_pretrained issues form the only clearly related family, while the rest of the items are plausibly separate bugs across different models/subsystems. The accepted soft-edge links are conservative enough given the titles, especially the identical 45310/45478 reports and the nearby 45375 config-field regression." }, "evaluator_used": true, "retried": false, @@ -14486,226 +14448,226 @@ "issue:43479|issue:44079", "issue:42898|issue:44779" ], - "prepared_review_unit_hash": "68eaee99381b9157347c995110b0aec7238fb5610902b9b1a877e8fa34f24b01", + "prepared_review_unit_hash": "8b2eb19ebb7773be25579094f23827bca6e54fd5e8c66b67b5c2d71866b3a63a", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37088, - "estimated_input_tokens": 9272, - "estimated_eval_tokens": 18800 + "serialized_chars": 37091, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous Transformers issues across tokenizers, generation, docs, quantization, and model-specific bugs. Only one clear duplicate pair stands out: the two Qwen3ForCausalLM VRAM-leak issues are the same runtime/threading bug.", + "summary": "Mostly unrelated Transformers issue reports; only one soft edge is an exact duplicate pair about Qwen3ForCausalLM VRAM leakage in multiple dataloader threads.", "confidence": 0.96, - "canonical_issue_reason": "issue:42673 is the strongest canonical issue: it is the earlier report of the exact Qwen3ForCausalLM multi-dataloader-thread VRAM leak, with more discussion and references.", + "canonical_issue_reason": "issue:42673 is the best canonical issue because it is the fuller, earlier report and exactly matches issue:43704 by title and bug description.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42673 best represents the cluster\u2019s only clear duplicate and has the most complete signal for that bug.", + "best_issue_reason": "issue:42673 is the strongest representative of the only true duplicate in this cluster; it has the more complete report and the duplicate later issue can be consolidated into it.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44857", "right": "issue:46877", "accept": false, - "reason": "Different bugs: FP16/CUDA loss crash vs docstring import-time logging noise." + "reason": "Different bugs: AMP/CUDA crash in LwDetrImageLoss vs auto_docstring printing diagnostics at import time." }, { "left": "issue:45081", "right": "issue:45103", "accept": false, - "reason": "Different crash sites and causes; both are AttributeError reports but in unrelated code paths." + "reason": "Both are AttributeError reports, but they affect different code paths: Mistral regex patching vs auto_docstring processing." }, { "left": "issue:42994", "right": "issue:43824", "accept": false, - "reason": "Quantized model saving failure is unrelated to the Qwen2.5-VL import error." + "reason": "Unrelated failures: quantized save_pretrained vs importing Qwen2_5_VLForConditionalGeneration." }, { "left": "issue:46202", "right": "issue:46701", "accept": false, - "reason": "Both mention Gemma, but one is token roundtrip failure and the other is a broken image-text-to-text pipeline." + "reason": "Both mention Gemma, but one is a tokenizer roundtrip issue and the other is an image-text-to-text generation pipeline failure." }, { "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Both involve MoE/FP8, but one is save_pretrained and the other is from_pretrained initialization; different failures." + "reason": "Both are MoE/FP8 related, but one is FP8 save_pretrained and the other is a from_pretrained config/load error." }, { "left": "issue:45593", "right": "issue:46877", "accept": false, - "reason": "D-FINE auxiliary-loss bug is unrelated to auto_docstring stderr/stdout diagnostics." + "reason": "Auxiliary-loss behavior in D-FINE is unrelated to auto_docstring import-time stdout noise." }, { "left": "issue:42673", "right": "issue:43704", "accept": true, - "reason": "Same exact issue: Qwen3ForCausalLM VRAM leak when used across multiple dataloader threads; titles and symptoms match." + "reason": "Exact duplicate: same title and same VRAM leak in Qwen3ForCausalLM when used across multiple dataloader threads." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation kwargs sync bug is unrelated to a Janus image-generation test failure." + "reason": "Different generation problems: sync_gpus kwargs handling vs janus image generation test failure." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "Both are FP8-related, but accuracy regression and FP8 save_pretrained failure are different underlying problems." + "reason": "Both involve FP8, but one is an accuracy concern and the other is a save_pretrained bug in MoE." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Model config import failure is unrelated to xdist race on captured_info.txt." + "reason": "Unrecognized model config error is unrelated to a pytest-xdist captured_info.txt race." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Package extra/backend issue is unrelated to a missing image_utils import." + "reason": "Backend packaging for TensorFlow vs a missing image_utils import are separate issues." }, { "left": "issue:44743", "right": "issue:47068", "accept": false, - "reason": "Recurrent-state/cache reset bug is unrelated to host synchronization in deepspeed loss computation." + "reason": "Recurrent state reset with cache is not the same as a potential host sync in deepspeed_sp_compute_loss." }, { "left": "issue:45561", "right": "issue:45600", "accept": false, - "reason": "xdist file race and auto_mappings stale config references are different CI issues." + "reason": "Test file race in captured_info.txt is unrelated to stale auto_mappings references causing CI failures." }, { "left": "issue:45405", "right": "issue:46877", "accept": false, - "reason": "Pinned dependency version mismatch is unrelated to doc-lint output at import time." + "reason": "Version pinning to unreleased PEFT is unrelated to auto_docstring import-time diagnostics." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version compatibility problem is unrelated to a missing module import." + "reason": "Python version compatibility and missing import are different failures." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLFlow reporting on failed training and Deepspeed OOM with Adafactor are different bugs." + "reason": "Both are training/runtime issues, but MLFlow reporting on failure and DeepSpeed Adafactor OOM are distinct bugs." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Stuck loading Sentence Transformers and embedding reinit on resize are unrelated." + "reason": "Sentence Transformers load hangs and token-embedding reinit on resize are unrelated." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Both tokenizer regressions in v5, but different behaviors and different tokenizer families." + "reason": "Both are tokenizer regressions in v5, but they describe different concrete behaviors and code paths." }, { "left": "issue:43479", "right": "issue:46829", "accept": false, - "reason": "Multimodal config default initialization bug is unrelated to incorrect hidden_states in VideoPrism output." + "reason": "Config initialization with None is unrelated to the VideoPrism output field mismatch." }, { "left": "issue:42175", "right": "issue:45712", "accept": false, - "reason": "Backend extra packaging and leftover dummy classes are separate maintenance issues." + "reason": "Package/backend inclusion and leftover dummy classes are separate maintenance issues." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Missing files in processor.save_pretrained is unrelated to Qwen3.5 MoE from_pretrained failure." + "reason": "Missing processor save_pretrained files is not the same bug as Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Gemma4 attention-mask bug and partial-batch pipeline termination bug are different code paths." + "reason": "Causal attention masks in Gemma4 and final-partial-batch handling in MaskGenerationPipeline are different code paths." }, { "left": "issue:45600", "right": "issue:47059", "accept": false, - "reason": "Stale auto_mappings config refs and huggingface_hub CLI incompatibility are unrelated CI failures." + "reason": "Auto_mappings references to removed configs and a huggingface_hub CLI completion error are unrelated." }, { "left": "issue:42673", "right": "issue:46693", "accept": false, - "reason": "Same model family, but VRAM leak in dataloader threads is not the same as FA2 long-context performance regression." + "reason": "VRAM leak in dataloader threads is not the same as the FA2 long-context performance regression." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Tokenizer loading failure and add_special_tokens/BOS-EOS regression are different tokenizer bugs." + "reason": "Tokenizer loading in v5 and add_special_tokens/BOS-EOS behavior are adjacent but not the same concrete bug." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Multimodal config initialization bug is unrelated to Whisper timestamp decoding crash." + "reason": "Multimodal config default initialization is unrelated to Whisper timestamp decoding crashing on a trailing replacement character." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer behavior regressions, but they affect different models and different failure modes." + "reason": "Both concern tokenizer regressions, but one is clean_up_tokenization_spaces behavior and the other is Granite token ID generation." }, { "left": "issue:44016", "right": "issue:46231", "accept": false, - "reason": "Notebook syntax error and docs parameter-name mismatch are unrelated documentation issues." + "reason": "Notebook syntax error and a docs mismatch on set_training parameter names are different documentation issues." }, { "left": "issue:39692", "right": "issue:46909", "accept": false, - "reason": "Both are broken docs examples, but they are different examples with different model-specific failures." + "reason": "Both are broken examples, but they involve different models and different failure modes." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Default config initialization and ModelOutput key assignment are unrelated internals." + "reason": "Config defaulting and ModelOutput key assignment are unrelated." }, { "left": "issue:42898", "right": "issue:44779", "accept": false, - "reason": "Both are v5 tokenizer regressions, but they concern different models and different incorrect outputs." + "reason": "Both are tokenizer regressions, but there is no evidence they are the same underlying bug or code fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft-edge pairs. The exact duplicate between issue:42673 and issue:43704 is clearly supported by the identical titles and matching symptom, and the canonical issue choice is reasonable given the earlier report. No overstatement stands out." + "feedback": "The summary is well grounded in the packet: the only clear duplicate pair is issue:42673 / issue:43704, and the other soft-edge rejections are conservative and consistent with the issue titles/descriptions." }, "evaluator_used": true, "retried": false, @@ -14721,8 +14683,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -14805,208 +14767,208 @@ "issue:41628|issue:42757", "issue:44297|issue:45310" ], - "prepared_review_unit_hash": "71004ef66e3a6b2215a386e9b7e9beb989f8e14279ac1f2daab59b3689b28aec", + "prepared_review_unit_hash": "ff3cd84b9553fc01066f0623d82da26cef852c545e87aad9ae25ba4c9a4ccdbb", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35727, - "estimated_input_tokens": 8932, - "estimated_eval_tokens": 18120 + "serialized_chars": 35724, + "estimated_input_tokens": 8931, + "estimated_eval_tokens": 18118 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is heterogeneous: the soft pairs are mostly related by broad topic only (imports, tokenizers, multimodal models, training, save/load paths), but none look like the same concrete bug or change. I would reject every soft edge and treat these as distinct issues.", - "confidence": 0.83, - "canonical_issue_reason": "No single cluster-wide canonical issue stands out, but issue 44297 is the strongest representative candidate because it is the most concrete, self-contained, and reproducible bug report in the set.", + "summary": "The candidate pairs are overwhelmingly unrelated; I would reject all soft edges. If a representative issue must be chosen, the clearest standalone regression report is the v4.57.5 processor/tokenizer breakage issue.", + "confidence": 0.95, + "canonical_issue_reason": "No strong duplicate cluster emerges from these issues; they span unrelated import errors, model-specific regressions, trainer bugs, and docs. If one representative issue is needed, issue 43295 is the most actionable because it describes a concrete regression with a clear failure path and affected version.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44297 is the best issue to anchor on: it has a specific failure mode and a narrow fix surface, unlike the broader or more ambiguous reports here.", + "best_issue_reason": "Issue 43295 is the best global representative: it is a specific, reproducible regression in a recent release, with a well-defined code path (processor.tokenizer / image passing) and clear user impact.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Different missing symbols and different dependency surfaces; these are both import errors, but not the same bug." + "reason": "Different import failures and different subsystems: sentence-transformers import vs missing Qwen2.5-VL class." }, { "left": "issue:27957", "right": "issue:30333", "accept": false, - "reason": "Flash Attention behavior for XLMRoberta is unrelated to failed-job MLflow reporting." + "reason": "Unrelated problems: Flash Attention behavior for XLM-Roberta vs MLflow reporting after failed training jobs." }, { "left": "issue:30333", "right": "issue:34689", "accept": false, - "reason": "One is about MLflow not closing a failed training run, the other is a model-loading regression." + "reason": "Different concrete bugs: failed-job MLflow tracking vs model loading regression for Llama 3.2 Vision." }, { "left": "issue:42898", "right": "issue:43295", "accept": false, - "reason": "Both involve v5 behavior changes, but one is tokenization cleanup and the other is processor/tokenizer image-passing regression." + "reason": "Both are v5 regressions, but one is tokenizer cleanup behavior and the other is processor/tokenizer access with images; not the same bug." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "Both are multimodal/model-setup problems, but they affect different models and different code paths." + "reason": "Different multimodal issues: vitpose model breakage vs Gemma 4 token-type defaults." }, { "left": "issue:43874", "right": "issue:45092", "accept": false, - "reason": "Missing an image-processor method is not the same as old remote-code checkpoints failing under meta initialization." + "reason": "Both are multimodal/v5 compatibility issues, but one is a missing image processor method and the other is remote-code meta-init incompatibility." }, { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "A buffer-read AttributeError and a double optimizer-step under deepspeed are distinct runtime bugs." + "reason": "Unrelated: position_ids buffer access error vs optimizer stepping twice under deepspeed." }, { "left": "issue:44734", "right": "issue:45103", "accept": false, - "reason": "Serving KV-cache indexing and auto_docstring annotation handling are unrelated failures." + "reason": "Serving KV-cache tensor indexing bug vs auto-docstring crash from future annotations; different code paths." }, { "left": "issue:42898", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but the specific symptoms and affected logic differ." + "reason": "Both involve tokenizers, but one is generic v5 cleanup behavior and the other is a Kimi-K2.5 codec regression; not the same underlying bug." }, { "left": "issue:30333", "right": "issue:36331", "accept": false, - "reason": "MLflow reporting on failure is unrelated to the Trainer compute_loss signature mismatch." + "reason": "Different training issues: MLflow status handling vs compute_loss signature incompatibility." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Importing transformers.onnx and missing PyTorch for AutoModel are different import/dependency problems." + "reason": "Both are import errors, but one is missing transformers.onnx and the other is missing PyTorch; separate causes." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "Both mention save_pretrained, but one is FP8/MoE serialization and the other is tokenizer metadata mismatch." + "reason": "Both mention save_pretrained, but one is FP8 MoE serialization and the other is tokenizer_class metadata mismatch." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "Optimizer argument handling and a documentation parameter-name mismatch are not the same issue." + "reason": "Optimizer behavior bug vs documentation typo in set_training; not the same issue." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "A version pin to an unreleased package and leftover dummy classes in the repo are unrelated." + "reason": "Unrelated: unreleased PEFT version pin vs leftover dummy classes leaking into dir(transformers)." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Both are example-script failures, but they involve different scripts and different bugs." + "reason": "Different example-script failures: multi-label classification returning empty results vs local dataset loading in image classification." }, { "left": "issue:44060", "right": "issue:45125", "accept": false, - "reason": "Both are Qwen3-related, but one is a tied-weights warning and the other is a missing tensor-parallel plan." + "reason": "Same family (Qwen), but different concrete failures: false tied-weights warning vs missing tensor-parallel plan." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FP8 accuracy concerns and missing processor save files do not share the same failure path." + "reason": "Different bugs: FineGrainedFP8 accuracy regression vs processor.save_pretrained missing files." }, { "left": "issue:44933", "right": "issue:45712", "accept": false, - "reason": "A missing image_utils import and dummy-class leakage into dir(transformers) are different export/surface issues." + "reason": "Both are import/cleanup related, but one is an image_utils import failure and the other is dummy class leakage; not the same defect." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to the deepspeed double-step optimizer bug." + "reason": "Tokenizer save metadata bug vs optimizer step-count bug; unrelated." }, { "left": "issue:44448", "right": "issue:46132", "accept": false, - "reason": "Pegasus output regression and AttentionInterface.register behavior change are not the same bug." + "reason": "Different concerns: Pegasus v4/v5 output change vs AttentionInterface.register behavior change." }, { "left": "issue:42796", "right": "issue:43723", "accept": false, - "reason": "Both are tokenizer-related, but one is BPE marker handling and the other is a general AutoTokenizer.from_pretrained loading issue." + "reason": "Both are tokenizer-related v5 reports, but one is a specific BPE-marker handling bug and the other is a broader AutoTokenizer.from_pretrained loading issue." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "A category-cardinality runtime error is unrelated to AttentionInterface registration semantics." + "reason": "Unrelated: category-cardinality runtime error vs attention registration behavior." }, { "left": "issue:43479", "right": "issue:43644", "accept": false, - "reason": "Default config initialization and non-persistent buffer junk are different code paths." + "reason": "Different code paths: multimodal config defaulting vs non-persistent buffer initialization." }, { "left": "issue:44222", "right": "issue:44623", "accept": false, - "reason": "FP8 MoE save_pretrained and processor.save_pretrained missing files are separate serialization issues." + "reason": "Both involve save_pretrained, but one is FP8 MoE serialization and the other is missing processor files." }, { "left": "issue:44360", "right": "issue:46762", "accept": false, - "reason": "A discussion about a DSA indexer ReLU and numerical differences in Minimax attention are not the same underlying bug." + "reason": "Different numerical/implementation issues: DSA indexer lacking ReLU vs Minimax MSA attention mismatch." }, { "left": "issue:46132", "right": "issue:46710", "accept": false, - "reason": "These are both performance/behavior regressions, but they affect different subsystems and symptoms." + "reason": "Different bugs: registered attention function behavior vs DeepSeek-R1-Distill-Llama accuracy regression." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "A wrong docstring is unrelated to the GLM-5 RoPE implementation bug." + "reason": "Docstring mismatch for position_ids vs GLM-5 RoPE implementation; not the same underlying bug." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but for different modules, symbols, and packages." + "reason": "Both are import errors, but one is a missing transformers symbol and the other is a huggingface_hub symbol mismatch." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch and Qwen3.5 MoE from_pretrained failure are different issues." + "reason": "Tokenizer save_pretrained metadata bug vs Qwen3.5 MoE from_pretrained regression; different operations and failure modes." } ] }, @@ -15025,8 +14987,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -15110,208 +15072,208 @@ "issue:44829|issue:46002", "issue:43408|issue:46402" ], - "prepared_review_unit_hash": "9259925ebed943deeaa7e59af73f93219d36a6846ccdd3fb6a52c6b66df1c876", + "prepared_review_unit_hash": "d69cf30596f4f277c6d77cc5928f6fbd2df6f370a6e0e77300e30289c01618b4", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36016, - "estimated_input_tokens": 9004, - "estimated_eval_tokens": 18264 + "serialized_chars": 36022, + "estimated_input_tokens": 9006, + "estimated_eval_tokens": 18268 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Broad, heterogeneous set of Transformers issues. The soft pairs mostly share only a subsystem or buzzwords (load/save, generation, video, quantization) and do not look like the same underlying bug or change.", + "summary": "The candidate set is a loose collection of Transformers bug reports with only superficial thematic overlap; none of the soft pairs look like true duplicates of the same underlying bug/change, so all soft edges are rejected.", "confidence": 0.92, - "canonical_issue_reason": "issue:44964 is the clearest high-signal representative of the cluster\u2019s recurring load/regression theme: a specific model fails to load on the latest Transformers version.", + "canonical_issue_reason": "issue:46305 is the clearest standalone bug report here: it names a precise off-by-one in a specific masking function, has an explicit expected-vs-actual behavior, and is concrete enough to serve as the best issue representative among otherwise unrelated reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44964 is the most actionable representative: it has a concrete failing model, a clear user-facing regression, and a narrow reproduction target.", + "best_issue_reason": "issue:46305 is the strongest issue to keep as the representative artifact because it is specific, reproducible, and narrowly scoped to one code path, unlike the other candidates which are either broader, more model-specific, or less directly actionable.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43504", "right": "issue:47059", "accept": false, - "reason": "Different failures: BEiT semantic-segmentation preset loading vs CLI completion breakage in huggingface_hub." + "reason": "Different failures in unrelated areas: pretrained model loading for BEiT vs CLI completion support in Hugging Face Hub integration." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "Tokenizer loading in v5 and Qwen3.5 flash-attention generation are unrelated code paths and symptoms." + "reason": "One is tokenizer loading in v5, the other is Qwen3.5 generation with flash-attention; different model paths and symptoms." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "PretrainedConfig type-checking regression is separate from LlamaConfig head_dim validation." + "reason": "Both involve config validation in v5, but the concrete bugs differ: PretrainedConfig type checking vs explicit head_dim rejection in LlamaConfig." }, { "left": "issue:43122", "right": "issue:46209", "accept": false, - "reason": "Tokenizer-version behavior change is unrelated to FP8 inference failure on B200." + "reason": "Tokenization regression and FP8 inference failure are unrelated bugs in different subsystems." }, { "left": "issue:46614", "right": "issue:46932", "accept": false, - "reason": "Stale phonemizer backend state is unrelated to Gemma4 quantized embedder dtype casting." + "reason": "Stale phonemizer language state is unrelated to Gemma4 quantized embedder dtype casting." }, { "left": "issue:44222", "right": "issue:44484", "accept": false, - "reason": "FP8 save_pretrained/moe bug is distinct from the max_shard_size default question." + "reason": "FP8 MoE save_pretrained failure is not the same as the max_shard_size default question." }, { "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "GlmMoeDsaIndexer split behavior is unrelated to InternVL2 meta-init incompatibility." + "reason": "GlmMoeDsaIndexer split-return bug is different from remote-code checkpoint incompatibility during meta initialization." }, { "left": "issue:44190", "right": "issue:46082", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to LlamaConfig head_dim rejection." + "reason": "Dataset loading in a training script is unrelated to LlamaConfig head_dim validation." }, { "left": "issue:43232", "right": "issue:46424", "accept": false, - "reason": "Generation kwargs after sync_gpus is a different bug from cache max_cache_length handling." + "reason": "Generation kwargs after sync_gpus and cache max length handling are different code paths with different behavior." }, { "left": "issue:46209", "right": "issue:46431", "accept": false, - "reason": "FP8 inference on B200 and processor.save_pretrained missing files are unrelated issues." + "reason": "FP8 inference on B200 and processor.save_pretrained missing files are unrelated." }, { "left": "issue:43232", "right": "issue:45834", "accept": false, - "reason": "Generation state handling is unrelated to Kosmos2.5 OCR index errors." + "reason": "Generation model-kwarg handling is unrelated to Kosmos2.5 OCR length indexing." }, { "left": "issue:44295", "right": "issue:44623", "accept": false, - "reason": "Reading position_ids after buffer registration is unrelated to processor.save_pretrained file omission." + "reason": "Position_ids buffer access and processor.save_pretrained file omission are different bugs." }, { "left": "issue:30990", "right": "issue:38175", "accept": false, - "reason": "SentenceTransformers loading stall and SIGLIP2 zero-probability outputs are different model/runtime problems." + "reason": "Sentence Transformers loading stalls and siglip2 zero probabilities are different model/runtime issues." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "A GlmMoeDsaIndexer split/return-value issue is not the same as a docstring typo for position_ids." + "reason": "A DSA indexer return-value bug is not the same as a wrong docstring for position_ids." }, { "left": "issue:46369", "right": "issue:46909", "accept": false, - "reason": "SAM3 LiteText memory overflow is unrelated to a broken documentation example." + "reason": "SAM3 memory overflow and a broken DIA documentation example are unrelated." }, { "left": "issue:46305", "right": "issue:46655", "accept": false, - "reason": "Both involve video/streaming processing, but one is a sliding-window off-by-one and the other is padding calculation in a different model path." + "reason": "Both touch video processing, but one is a sliding-window off-by-one and the other is a padding-calculation bug." }, { "left": "issue:43475", "right": "issue:46305", "accept": false, - "reason": "Missing fpn_position_embeddings in SAM3 output is unrelated to MoonshineStreaming window masking." + "reason": "SAM3 encoder attribute missing and MoonshineStreaming mask off-by-one are different model/components." }, { "left": "issue:44263", "right": "issue:44360", "accept": false, - "reason": "Same component family, but different concrete bugs: torch.split return handling vs missing ReLU in the DSA indexer." + "reason": "Same broad DSA indexer area, but one reports split-return semantics and the other a missing ReLU; not the same bug." }, { "left": "issue:42907", "right": "issue:46050", "accept": false, - "reason": "Dequantized Ministrals/Devstrals save failure and sinq quantization save/load breakage are different model-specific bugs." + "reason": "Both mention save/load of quantized models, but the affected mechanisms and failure modes differ." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "FineGrainedFP8 accuracy issues are unrelated to Qwen3.5 MoE from_pretrained errors." + "reason": "FineGrainedFP8 accuracy issues and Qwen3.5 from_pretrained errors are unrelated." }, { "left": "issue:45092", "right": "issue:45357", "accept": false, - "reason": "Meta-initialization incompatibility for old InternVL2 checkpoints is separate from incorrect visual encoder keys on save_pretrained." + "reason": "Meta-init incompatibility for old remote-code checkpoints is not the same as incorrect visual encoder keys saved by Qwen3.5." }, { "left": "issue:44279", "right": "issue:46328", "accept": false, - "reason": "Generic dependency trouble is unrelated to SequenceFeatureExtractor.pad list-conversion overhead." + "reason": "A generic dependency issue is unrelated to SequenceFeatureExtractor.pad inefficiency." }, { "left": "issue:22355", "right": "issue:27957", "accept": false, - "reason": "Missing transformers.onnx module and XLMRoberta Flash Attention 2 are unrelated issues." + "reason": "Missing transformers.onnx module and XLMRoberta Flash Attention 2 behavior are unrelated issues." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "Chat completions request rejection is unrelated to Gemma4Processor missing _tokenizer." + "reason": "Chat completions request rejection and Gemma4Processor missing _tokenizer are different server/processor failures." }, { "left": "issue:44964", "right": "issue:47059", "accept": false, - "reason": "Phi-4 multimodal loading regression is unrelated to the CLI completion attribute error in huggingface_hub." + "reason": "Phi-4 multimodal loading and CLI completion breakage are unrelated." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "FP8 accuracy regression and Qwen3-VL-MoE loading trouble are different problems." + "reason": "FP8 accuracy regression and Qwen3-VL-Moe v5 bug are different model-specific issues." }, { "left": "issue:45381", "right": "issue:46305", "accept": false, - "reason": "Qwen2.5-VL vision_position_ids mismatch is not the same bug as MoonshineStreaming mask off-by-one." + "reason": "Both concern vision/video positions, but one is qwen2.5-vl vision_position_ids and the other is MoonshineStreaming masking; not the same code path." }, { "left": "issue:44829", "right": "issue:46002", "accept": false, - "reason": "Flash-Attention-3 degenerate training and NaNs after checkout+init are distinct training stability issues." + "reason": "Flash-attention-3 degenerate training and NaNs in classification heads are different training stability bugs." }, { "left": "issue:43408", "right": "issue:46402", "accept": false, - "reason": "sam3_video/sam3_tracker type warning is unrelated to SAM3 text encoder weights not loading." + "reason": "sam3_tracker/sam3_video model-type mismatch is unrelated to SAM3 text encoder weights not loading." } ] }, @@ -15330,8 +15292,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -15399,8 +15361,8 @@ "issue:27957|issue:30990", "issue:42796|issue:43122", "issue:45381|issue:46909", - "issue:43540|issue:46909", "issue:43232|issue:45593", + "issue:43540|issue:46909", "issue:43646|issue:46002", "issue:42907|issue:44945", "issue:30990|issue:33357", @@ -15413,202 +15375,202 @@ "issue:39401|issue:44488", "issue:41628|issue:45278" ], - "prepared_review_unit_hash": "b16f49d3bdfda4bb1ddaf283fbce322e2ca988fc75bbf76dd62b956a44373bc2", + "prepared_review_unit_hash": "322cedd2cc06ce15b88b64252e7bbeef01660da489df270cf5d0127bf4a38cf3", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35157, - "estimated_input_tokens": 8790, - "estimated_eval_tokens": 17836 + "serialized_chars": 35153, + "estimated_input_tokens": 8789, + "estimated_eval_tokens": 17834 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are a mixed bag of unrelated Transformers issues (imports, tokenizers, multimodal/video, quantization, and v5 migration regressions). None of the soft pairs look like true duplicates; they mostly share only broad theme words.", - "confidence": 0.91, - "canonical_issue_reason": "issue:45278 is the closest umbrella candidate because it covers the broadest import/API regression theme, but even that is only a loose thematic match rather than a true duplicate target.", + "summary": "The items are mostly unrelated regression reports that only look similar at the title level. None of the soft-similarity pairs are strong enough to merge as duplicates. If a single issue must be chosen as representative, issue 42886 is the clearest standalone bug report, but there is no true cluster-wide canonical.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 42886 is the most concrete, reproducible bug report in the set (version-specific offline cache failure), so it is the least ambiguous representative if one issue must be picked.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single representative to keep visible because it is the broadest user-facing regression report in the set, especially for upgrade-related import failures.", + "best_issue_reason": "Issue 42886 is the strongest standalone candidate: it has a precise failure mode, a pinned version, and an actionable reproduction path. The rest of the set is too heterogeneous to serve as a shared canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41628", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but for different missing symbols and likely different export/code paths." + "reason": "Both are import errors, but they concern different symbols and different failure contexts; not the same bug." }, { "left": "issue:44261", "right": "issue:45092", "accept": false, - "reason": "Both involve v5/meta-init migration, but one is an MLA layernorm precision/config bug and the other is remote-code checkpoint incompatibility." + "reason": "One is a q_a_layernorm epsilon precision issue, the other is a Transformers v5 meta-init incompatibility for old InternVL2 checkpoints." }, { "left": "issue:43644", "right": "issue:46305", "accept": false, - "reason": "Non-persistent buffer junk vs sliding-window off-by-one are unrelated bugs in different code paths." + "reason": "Different code paths: non-persistent buffer corruption vs MoonshineStreaming sliding-window off-by-one." }, { "left": "issue:42831", "right": "issue:45753", "accept": false, - "reason": "FP8 accuracy regression and an image-triggered infinite loop are different failure modes." + "reason": "One is an FP8 accuracy regression; the other is an infinite-loop bug on a specific Qwen3_5 image input." }, { "left": "issue:44327", "right": "issue:47059", "accept": false, - "reason": "QA span-decoding top-k crash and CLI completion attribute error are unrelated subsystems." + "reason": "QA pipeline top-k bounds crash and CLI completion attribute error are unrelated." }, { "left": "issue:22355", "right": "issue:28218", "accept": false, - "reason": "Missing module import and tokenizer spacing behavior are not the same issue." + "reason": "Module import failure and tokenizer spacing behavior are not the same underlying bug." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer bugs, but they affect different models and different outputs." + "reason": "Both are tokenizer-related, but one is offset_mapping correctness and the other is add_special_tokens/BOS-EOS behavior for a specific model." }, { "left": "issue:42886", "right": "issue:42994", "accept": false, - "reason": "Offline cache loading and quantized model saving are separate tokenizer vs serialization problems." + "reason": "Offline cache loading and quantized model saving are different lifecycle paths with different failure modes." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "Attention-kernel divergence under compile and a docstring typo are not the same bug." + "reason": "Torch.compile attention-kernel divergence is unrelated to a docstring issue." }, { "left": "issue:42886", "right": "issue:43066", "accept": false, - "reason": "Offline tokenizer loading failure and wrong decoder type in v5 are different code paths." + "reason": "Offline tokenizer cache loading and wrong tokenizer decoder type in v5 are distinct regressions." }, { "left": "issue:42757", "right": "issue:45278", "accept": false, - "reason": "A specific huggingface_hub import failure is not the same as a broad Transformers v5 import-error report." + "reason": "The second is a broad multi-import regression, but it is not the same concrete import symbol breakage as the first." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "sam3 model-type warning and sam3-litetext memory overflow are unrelated loading vs memory issues." + "reason": "Both mention sam3, but one is a model-type warning and the other is a memory overflow; different bugs." }, { "left": "issue:27957", "right": "issue:30990", "accept": false, - "reason": "Flash Attention 2 on XLM-Roberta and Sentence Transformers loading stall are different behaviors." + "reason": "Flash Attention 2 support for XLMRoberta and Sentence Transformers loading hangs are unrelated." }, { "left": "issue:42796", "right": "issue:43122", "accept": false, - "reason": "Model-specific BPE marker handling differs from a broader version-to-version tokenization change." + "reason": "Both involve tokenization, but one is BPE marker handling for a specific model and the other is a version-to-version tokenization change." }, { "left": "issue:45381", "right": "issue:46909", "accept": false, - "reason": "Qwen2.5-VL video position IDs and a DIA docs example are unrelated." + "reason": "Video position_ids correctness and a broken documentation example are not the same issue." }, { - "left": "issue:43540", - "right": "issue:46909", + "left": "issue:43232", + "right": "issue:45593", "accept": false, - "reason": "Video-input processing in Qwen3OmniMoe and a broken documentation example are unrelated." + "reason": "Generation kwargs handling after sync_gpus and D-FINE auxiliary-loss logic are unrelated." }, { - "left": "issue:43232", - "right": "issue:45593", + "left": "issue:43540", + "right": "issue:46909", "accept": false, - "reason": "Generation kwargs after sync_gpus and D-FINE auxiliary-loss logic are unrelated." + "reason": "Qwen3OmniMoe video-input processing and a DIA documentation example are unrelated." }, { "left": "issue:43646", "right": "issue:46002", "accept": false, - "reason": "Custom model init breakage and NaNs after checkout/init are different initialization issues." + "reason": "Custom model initialization breakage and NaNs after checkout/init are different failure modes." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Quantized-model save failure and pipeline-parallelism output corruption are unrelated." + "reason": "Quantized model saving failure and pipeline-parallelism output corruption are separate bugs." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Loading stall for sentence-transformers and a MacOS CLIP bus error are different model/runtime failures." + "reason": "Both are runtime/loading problems, but they affect different models, platforms, and symptoms." }, { "left": "issue:44295", "right": "issue:44297", "accept": false, - "reason": "position_ids buffer access and tokenizer_class metadata mismatch are different tokenizer/model issues." + "reason": "A buffer read error and tokenizer_config class mismatch are different serialization/loading issues." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python version compatibility and pipeline-parallel output errors are unrelated." + "reason": "Python version compatibility and incorrect LLM output under pipeline parallelism do not share a code path." }, { "left": "issue:43329", "right": "issue:47059", "accept": false, - "reason": "Multimodal token counting and CLI group completion are different areas." + "reason": "Multimodal token counting bug and CLI completion attribute error are unrelated." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Saving dequantized models and loading Phi-4 multimodal are different bugs." + "reason": "Saving dequantized Ministrals/Devstrals and loading Phi-4 multimodal are different model-loading/saving issues." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "CLI env command breakage and sentence-transformers import NameError are different failure surfaces." + "reason": "A broken transformers env command and a sentence-transformers import NameError are unrelated." }, { "left": "issue:43756", "right": "issue:44060", "accept": false, - "reason": "RoPE-layer dropping and tied-weights warnings are different model-architecture bugs." + "reason": "Smollm3 RoPE-layer handling and Qwen3-Next tied-weights warning are different model-specific bugs." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Tokenizer offset mapping and a model-specific load failure are not duplicates." + "reason": "Wrong offset_mapping and inability to load cjvt/sleng-bert are not the same underlying tokenizer bug." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "A specific AutoImageProcessor import error and a broad import-error regression are only thematically related." + "reason": "A specific missing export and a broad set of import errors after upgrade are related only at a very high level, not as one duplicate bug." } ] }, @@ -15627,8 +15589,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -15711,20 +15673,20 @@ "issue:43493|issue:45753", "issue:44188|issue:46328" ], - "prepared_review_unit_hash": "130eb9cb8615bc2739f16a5cc132ca35b8b163855e68b2983a2e7ac49ca825c3", + "prepared_review_unit_hash": "a240b198c28f7e3caff2fa2f2ee3c194d32ba723acd690cc6e400aeb1f74d283", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35536, + "serialized_chars": 35534, "estimated_input_tokens": 8884, "estimated_eval_tokens": 18024 }, @@ -15733,186 +15695,186 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a noisy mixed cluster spanning tokenizer regressions, loading/import issues, save/pretrained bugs, and unrelated model-specific failures. None of the soft pairs are close enough to be treated as the same underlying bug/change.", - "confidence": 0.78, - "canonical_issue_reason": "issue:43122 is the broadest and most central tokenizer-regression report in the set; several nearby issues are also tokenizer-related, but this one best captures the shared failure mode without being too narrow.", + "summary": "This cluster is heterogeneous: the items span unrelated tokenizer, model-loading, training, packaging, multimodal, and parallelism bugs. None of the soft-similarity pairs look like the same underlying defect, so I would not merge any of them into a duplicate cluster.", + "confidence": 0.89, + "canonical_issue_reason": "No canonical issue: the cluster is heterogeneous and none of the items appear to describe the same underlying bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43122 is the best representative issue because it sits at the center of the tokenizer behavior regressions, which are the strongest recurring theme among the candidates.", + "best_issue_reason": "No single issue is a strong global representative; the set spans unrelated subsystems and model families.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42371", "right": "issue:44263", "accept": false, - "reason": "TF32 backend settings vs a GlmMoeDsaIndexer torch.split return-value bug are unrelated problems." + "reason": "TF32 backend settings and a GlmMoeDsaIndexer split bug are unrelated." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "Both involve from_pretrained/loading, but one is a TOKENIZER_MAPPING_NAMES fallback issue and the other is cached_file kwarg forwarding; different code paths and symptoms." + "reason": "TOKENIZER_MAPPING_NAMES returning None and AutoProcessor kwargs forwarding are different code paths." }, { "left": "issue:42907", "right": "issue:43937", "accept": false, - "reason": "A dequantized save_pretrained failure and an invalid GenerationConfig error are separate model-serialization vs generation-validation bugs." + "reason": "Saving dequantized models and invalid GenerationConfig handling are distinct failures." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Wrong offset_mapping in Qwen3 tokenizer and all-zero assistant_masks in multimodal chat templating are different tokenizer features and failure modes." + "reason": "Qwen3 offset_mapping bugs and multimodal assistant mask generation are not the same issue." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "A sentence-transformers import NameError and stale auto_mappings references to removed Sam3LiteText configs are unrelated import/config issues." + "reason": "An import NameError in sentence-transformers and removed Sam3LiteText configs are unrelated." }, { "left": "issue:46172", "right": "issue:46762", "accept": false, - "reason": "Both concern attention math, but they target different models/implementations and distinct numerical mismatches." + "reason": "Deepseek-v4 dtype/transpose behavior and Minimax numerical mismatch are different model-specific problems." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "Qwen3.5 num_labels propagation and flash_attention_3 degenerate training are different bugs with no shared code-path evidence." + "reason": "num_labels propagation and flash_attention_3 degenerate training do not share the same root cause." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "cached_file kwargs omission and apply_chat_template crashing on tool-call assistant messages are different APIs and failure conditions." + "reason": "cached_file kwarg forwarding and apply_chat_template tool-call crashes affect different paths." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "A missing ReLU in a DSA indexer is not the same as a GLM-5 RoPE implementation discrepancy." + "reason": "A missing ReLU in the DSA indexer is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Tokenizer version mismatch/regression and multimodal assistant mask generation are different behaviors." + "reason": "Tokenizer version drift and multimodal assistant mask bugs are not the same defect." }, { "left": "issue:43122", "right": "issue:46326", "accept": false, - "reason": "General tokenizer output drift and continue_final_message behavior for mistral_common tokenizers are distinct chat-templating/tokenizer issues." + "reason": "Both involve tokenizers, but one is a version regression and the other is a continue_final_message bug in mistral_common tokenizers." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "A model-type mismatch warning and a missing fpn_position_embeddings attribute are separate SAM 3 video issues." + "reason": "A model-type warning and a missing fpn_position_embeddings attribute are different SAM3 issues." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "Qwen3-VL-Moe problems in transformers 5.0 and tokenizer_class mismatch on save_pretrained are different failure surfaces." + "reason": "Qwen3-VL-Moe loading and tokenizer_class mismatch on save_pretrained are separate problems." }, { "left": "issue:42907", "right": "issue:43452", "accept": false, - "reason": "Saving dequantized models and gguf_file breaking from_pretrained loading are unrelated persistence/load bugs." + "reason": "Dequantized save failures and gguf_file from_pretrained breakage are not the same bug." }, { "left": "issue:22355", "right": "issue:29159", "accept": false, - "reason": "A missing transformers.onnx module and slow-vs-fast tokenizer inconsistency are separate historical issues." + "reason": "Missing transformers.onnx and slow/fast tokenizer inconsistency are unrelated." }, { "left": "issue:44038", "right": "issue:44295", "accept": false, - "reason": "Qwen3-VL-Moe bug and position_ids buffer read error do not appear to share a concrete code path." + "reason": "Qwen3-VL-Moe loading issues and position_ids buffer access are different code paths." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "A save_pretrained regression for Qwen3.5 and a from_pretrained error for Qwen3.5 MoE are related by model family but not clearly the same bug." + "reason": "Both are Qwen3.5-related, but one is save_pretrained corruption and the other is from_pretrained loading error; not enough to treat as one bug." }, { "left": "issue:44038", "right": "issue:44222", "accept": false, - "reason": "Qwen3-VL-Moe behavior and FP8 save_pretrained for moe are both MoE-adjacent but describe different bugs." + "reason": "Qwen3-VL-Moe model loading and FP8 moe save_pretrained issues are distinct." }, { "left": "issue:39401", "right": "issue:46614", "accept": false, - "reason": "Wrong offset mapping in Qwen3 tokenizer and stale phonemizer language state in Wav2Vec2PhonemeCTCTokenizer are unrelated tokenization subsystems." + "reason": "Qwen3 offset_mapping and Wav2Vec2 phonemizer language-state bugs are unrelated." }, { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "Missing Qwen2_5_VLForConditionalGeneration export and missing HybridCache export are separate import/export regressions." + "reason": "Missing Qwen2_5_VLForConditionalGeneration and missing HybridCache imports are different import regressions." }, { "left": "issue:43116", "right": "issue:45200", "accept": false, - "reason": "Empty multi-label classification results in an example script and Gemma 4 mm_token_type_ids defaults are different application-layer issues." + "reason": "Example-script multi-label output failure and Gemma 4 mm_token_type_ids defaults are different symptoms." }, { "left": "issue:43122", "right": "issue:44779", "accept": false, - "reason": "General tokenization drift in v5 and Deepseek-specific incorrect tokenization are related only at a very high level, not the same bug." + "reason": "A tokenizer version regression and a DeepSeek v5 tokenizer mismatch are not clearly the same root cause." }, { "left": "issue:42757", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols from different packages and different causes." + "reason": "Hugging Face Hub import failure and transformers HybridCache import failure are separate dependency issues." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training bug and empty multi-label classification output are unrelated training vs example-script issues." + "reason": "Tensor parallel training bugs and a classification example script returning empty results are unrelated." }, { "left": "issue:43278", "right": "issue:44945", "accept": false, - "reason": "Embedding dtype changing between train/eval and incorrect LLM output under pipeline parallelism are different distributed-training issues." + "reason": "BF16-to-FP32 dtype drift and pipeline parallelism output corruption are different training/runtime issues." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "Missing transformers.onnx and an unclear LayoutLMv3 error message are unrelated bugs." + "reason": "An import error for transformers.onnx and a LayoutLMv3 error-message clarity issue are unrelated." }, { "left": "issue:42175", "right": "issue:44938", "accept": false, - "reason": "TensorFlow missing from a pip extras install and Python 3.14 import/load failure are separate environment/package issues." + "reason": "Missing tensorflow in the torch extra and Python 3.14 import failure are unrelated packaging/runtime issues." }, { "left": "issue:43493", "right": "issue:45753", "accept": false, - "reason": "SigLIP2 implementation discrepancy and a Qwen3.5 infinite loop on a specific image are different model-specific failures." + "reason": "SigLIP2 implementation mismatch and Qwen3_5 infinite-loop behavior are not the same bug." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "torch.compile attention-kernel divergence and SequenceFeatureExtractor.pad inefficiency are unrelated performance/correctness issues." + "reason": "torch.compile attention-kernel divergence and a slow padding inefficiency do not share a concrete code-path failure." } ] }, @@ -15931,8 +15893,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -15991,8 +15953,8 @@ "issue:43937|issue:44485", "issue:38617|issue:45278", "issue:43066|issue:44779", - "issue:43611|issue:46002", "issue:44038|issue:45478", + "issue:43611|issue:46002", "issue:43976|issue:45446", "issue:42673|issue:45072", "issue:44355|issue:46231", @@ -16015,202 +15977,202 @@ "issue:45092|issue:46762", "issue:33453|issue:39401" ], - "prepared_review_unit_hash": "472a3f70b021eab39d161d0f0e56b5d27d6b3b3174ecb38d4c3e661418d64f59", + "prepared_review_unit_hash": "e64b89ad85fc540dbdbb0444a991304a935d4c29a3f48277303331d7e38534c7", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35738, - "estimated_input_tokens": 8935, - "estimated_eval_tokens": 18126 + "serialized_chars": 35735, + "estimated_input_tokens": 8934, + "estimated_eval_tokens": 18124 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The similarity pairs are mostly loose topical overlaps rather than true duplicates; they span unrelated bugs in tokenizers, model loading, processors, docs, and v5 regressions. I would not merge any of the soft-edge pairs.", - "confidence": 0.92, - "canonical_issue_reason": "issue:33453 is the broadest, most generic regression report in the set, but the cluster is too heterogeneous to have a strong canonical issue.", + "summary": "The items are a loose collection of unrelated bug reports; none of the soft pairs look like true duplicates or the same fixable code-path issue. Most similarities are only at the model/tokenizer/Transformer-v5 theme level, not the concrete underlying bug.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:33453 is the best umbrella candidate only because its title is the most general; it is not a true representative of the whole set.", + "best_issue_reason": "issue:46614 is the strongest standalone candidate: it is open, clearly described, has inbound references, and reports a specific reproducible regression in Wav2Vec2PhonemeCTCTokenizer state handling.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46231", "right": "issue:46877", "accept": false, - "reason": "Both are doc-related, but one is a parameter-name doc mismatch and the other is an import-time lint output problem." + "reason": "Docs parameter-name mismatch vs import-time docstring lint output; unrelated problems." }, { "left": "issue:42898", "right": "issue:46614", "accept": false, - "reason": "Both involve tokenizer behavior, but one is clean_up_tokenization_spaces semantics and the other is stale phonemizer language state." + "reason": "v5 tokenization-space behavior change vs stale espeak backend language state; different tokenizer subsystems and bugs." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "Same GLM-5 area, but the bugs are different: GenerationConfig validation vs RoPE implementation." + "reason": "GenerationConfig validation error vs RoPE implementation discussion; not the same defect." }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "45278 is a broad import-error umbrella; 38617 is a specific missing-symbol ImportError, so they are not the same bug." + "reason": "Single ImportError for one missing symbol vs many import errors after upgrade; too broad and not the same root cause." }, { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions in v5, but the affected tokenizer and concrete failure mode are different." + "reason": "Wrong decoder type in v5 vs DeepSeek tokenizer incorrect results; both tokenizer-related but different concrete failures." }, { - "left": "issue:43611", - "right": "issue:46002", + "left": "issue:44038", + "right": "issue:45478", "accept": false, - "reason": "Loading with base_model_prefix and NaNs in classification heads are unrelated code paths." + "reason": "Qwen3-VL-Moe bug report vs Qwen3.5 Moe from_pretrained error; same family, different code paths and symptoms." }, { - "left": "issue:44038", - "right": "issue:45478", + "left": "issue:43611", + "right": "issue:46002", "accept": false, - "reason": "Both mention Qwen3 Moe and v5, but they target different model variants and different load failures." + "reason": "base_model_prefix loading regression vs NaNs in classification heads; different model-loading versus initialization issues." }, { "left": "issue:43976", "right": "issue:45446", "accept": false, - "reason": "Python version compatibility is unrelated to the flex_attention PyTorch version check bug." + "reason": "Python version compatibility regression vs incorrect PyTorch version check in flex_attention; unrelated version-gating bugs." }, { "left": "issue:42673", "right": "issue:45072", "accept": false, - "reason": "A VRAM leak in threaded Qwen3 inference is not the same as a bfloat16 dtype mismatch in other models." + "reason": "VRAM leak in dataloader threads vs dtype mismatch in bfloat16 inference; distinct runtime problems." }, { "left": "issue:44355", "right": "issue:46231", "accept": false, - "reason": "Compiled-file execution errors are unrelated to a training API documentation typo." + "reason": "Compiled Python file errors vs documentation parameter-name typo; not the same bug." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Save/load serialization of visual encoder keys and from_pretrained load failures are distinct code paths." + "reason": "save_pretrained visual encoder key regression vs Qwen3.5 from_pretrained failure; related model family but different serialization/loading defects." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "num_labels propagation is a config bug; NaNs in classification heads are a separate model behavior issue." + "reason": "num_labels propagation bug vs NaNs in classification heads; different mechanisms and fixes." }, { "left": "issue:43116", "right": "issue:43493", "accept": false, - "reason": "An example-script multi-label bug and a SigLIP2 implementation discrepancy are not the same underlying problem." + "reason": "Empty multi-label example output vs SigLIP2 HF/JAX discrepancy; separate application-level and model-implementation issues." }, { "left": "issue:44610", "right": "issue:46909", "accept": false, - "reason": "Processor image-size padding is unrelated to a broken documentation example." + "reason": "Model input-size mismatch vs broken documentation example; one is code behavior, the other docs." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "The Qwen3 Omni feature-length helper bug and the flex_attention import check bug are unrelated." + "reason": "Helper output-length calculation bug vs flex_attention import version check; unrelated components." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 API guidance and a missing image-patch helper are different issues." + "reason": "TF32 API-setting guidance vs missing image-patch helper in GLM46V; not the same issue." }, { "left": "issue:46726", "right": "issue:46981", "accept": false, - "reason": "One is an AutoProcessor backend-missing error; the other is repeated HTTP 429 retry behavior." + "reason": "Missing vision backend handling vs repeated 429 retries in AutoProcessor; both processor-related but different failure modes." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "SGD argument handling and trainer evaluate-after-train failure are different failures." + "reason": "SGD args ignored vs trainer.evaluate() failing after train(); distinct training bugs." }, { "left": "issue:43618", "right": "issue:46829", "accept": false, - "reason": "Both concern output dataclasses, but they affect different models and different fields." + "reason": "CLIPOutput attentions missing vs VideoPrism output field mis-assignment; different model outputs." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "from_pretrained missing a weight vs AutoProcessor not forwarding kwargs to cached_file are separate bugs." + "reason": "Missing embed_tokens.weight in UMT5Encoder vs AutoProcessor not forwarding kwargs; unrelated loading/dispatch bugs." }, { "left": "issue:43066", "right": "issue:46710", "accept": false, - "reason": "A tokenizer decoder-type regression is not the same as a DeepSeek accuracy regression." + "reason": "Wrong tokenizer decoder type vs DeepSeek accuracy regression after upgrade; not the same underlying problem." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "The example-script classification bug and the pipeline translation warning are different problems." + "reason": "Example script returning empty results vs pipeline translation-task error message; different user-facing behaviors." }, { "left": "issue:28348", "right": "issue:46640", "accept": false, - "reason": "Both mention Flash Attention 2, but GPT2 support and a separate FA2 issue are not the same concrete change." + "reason": "GPT2 flash-attention support request vs Flash Attention 2 issue; same broad topic only, but no concrete same bug/change evidence." }, { "left": "issue:46614", "right": "issue:46752", "accept": false, - "reason": "Stale phonemizer language state is unrelated to an empty-list crash in apply_chat_template." + "reason": "Stale phonemizer backend state vs apply_chat_template crashing on empty list; unrelated tokenizer edge cases." }, { "left": "issue:44995", "right": "issue:46697", "accept": false, - "reason": "A stale cache crash on second forward pass is unrelated to over-padding in an image processor." + "reason": "Second-forward stale indexer cache crash vs image processor over-padding; different model and preprocessing bugs." }, { "left": "issue:42673", "right": "issue:42796", "accept": false, - "reason": "Different models and different bugs: VRAM leak vs tokenizer BPE marker handling." + "reason": "VRAM leak in multi-threaded inference vs tokenizer BPE-marker handling; not the same defect." }, { "left": "issue:45092", "right": "issue:46762", "accept": false, - "reason": "Meta-initialization incompatibility and numerical attention discrepancies are not the same issue." + "reason": "Remote-code/meta-init incompatibility vs numerical differences in Minimax attention; separate model-loading versus math-implementation issues." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Generic tokenizer-loading regression and a Qwen3 offset_mapping bug are too different to be duplicates." + "reason": "Tokenizer loading regression vs Qwen3 offset_mapping bug; both tokenizer-related but different root causes and scopes." } ] }, @@ -16229,8 +16191,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -16315,208 +16277,214 @@ "issue:38617|issue:41628", "issue:43122|issue:46298" ], - "prepared_review_unit_hash": "6a36422946b0f256787b95737f6c48a93dbeb2fa7fc89d474890ffb270fda1a4", + "prepared_review_unit_hash": "0946d85c9a971751dffae81cefe86fc914005b624ed703ebdf0abbcabe16bda2", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36719, - "estimated_input_tokens": 9180, - "estimated_eval_tokens": 18616 + "serialized_chars": 36732, + "estimated_input_tokens": 9183, + "estimated_eval_tokens": 18622 }, "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; I did not find any soft pair that looks like the same underlying defect, so no merges are justified.", - "confidence": 0.33, - "canonical_issue_reason": "If a single issue must serve as the umbrella, #43122 is the broadest version-regression/tokenization report and has the most general framing among the set.", + "summary": "The set is a heterogeneous mix of unrelated Transformers issues (tokenization, model loading, attention backends, training, processors, testing utils, and CLI/environment regressions). No soft pair looks like the same underlying bug or change, so all proposed duplicate links are rejected.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43122 is the broadest, most representative regression report in the set: a version-to-version tokenization change with cross-version comparison and inbound references, making it the best single issue anchor among otherwise diverse reports.", "canonical_pr_reason": null, - "best_issue_reason": "#43122 is the best triage anchor because it is a reusable cross-version tokenizer regression report, whereas most others are narrow model-specific or subsystem-specific bugs.", + "best_issue_reason": "issue:43122 is the most generally useful issue for the cluster because it captures a concrete v5 regression in core tokenizer behavior rather than a model-specific or peripheral failure.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44263", "right": "issue:46762", "accept": false, - "reason": "Different failures: a GlmMoeDsaIndexer split/indexing issue vs a Minimax MSA numerical discrepancy." + "reason": "Different bugs in different code paths: GlmMoeDsaIndexer split handling vs Minimax MSA attention numerical mismatch." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing pipeline() after v5 removals is unrelated to a PyTorch version check bug in flex_attention." + "reason": "Docs still referencing pipeline() is unrelated to the flex_attention AuxRequest version-check import bug." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "FSDP2/PEFT optimizer-state corruption and missing generation cache during chunked prefill are different code paths." + "reason": "FSDP2/PEFT optimizer state corruption and GenerationMixin cache loss are separate training vs generation failures." }, { "left": "issue:42222", "right": "issue:43441", "accept": false, - "reason": "vitpose model breakage and Ministral FlashAttention failure are unrelated model/backend bugs." + "reason": "Vitpose model breakage and Ministral flash-attention failure affect different models and subsystems." }, { "left": "issue:43824", "right": "issue:46981", "accept": false, - "reason": "Import error for Qwen2.5-VL is unrelated to repeated HTTP 429 retries in AutoProcessor." + "reason": "Qwen2.5-VL import error is unrelated to AutoProcessor retry behavior on HTTP 429." }, { "left": "issue:43673", "right": "issue:46328", "accept": false, - "reason": "Generation cache missing vs a padding performance inefficiency are clearly different issues." + "reason": "Cache handling in generation and numpy-to-list overhead in feature padding are unrelated." }, { "left": "issue:42222", "right": "issue:47000", "accept": false, - "reason": "vitpose loading breakage does not match RTDetrLoss infeasible cost-matrix failures." + "reason": "Vitpose model loading and RTDetr loss infeasibility are different issues." }, { "left": "issue:42673", "right": "issue:46655", "accept": false, - "reason": "VRAM leak in multi-threaded Qwen3ForCausalLM is unrelated to a SmolVLM video-padding bug." + "reason": "VRAM leakage in Qwen3ForCausalLM threads is unrelated to SmolVLM video padding shape errors." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "Invalid GenerationConfig and a stale warning for num_return_sequences are different generation-layer problems." + "reason": "Invalid GenerationConfig and stale num_return_sequences warning are distinct generation-config problems." }, { "left": "issue:42222", "right": "issue:46752", "accept": false, - "reason": "vitpose model breakage is unrelated to apply_chat_template crashing on an empty list." + "reason": "Vitpose model breakage is unrelated to empty-list chat template crashes." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and flash_attention crashing on s_aux=None are different mechanisms and subsystems." + "reason": "TOKENIZER_MAPPING_NAMES returning None and flash_attention s_aux=None crashes are separate failures." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "A PEFT version pin issue is unrelated to the flex_attention PyTorch version gate." + "reason": "An unreleased PEFT version bump is unrelated to the flex_attention import version check." }, { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "init_empty_weights argument mismatch and PretrainedConfig type-checking regression are distinct breakages." + "reason": "Unexpected _is_hf_initialized argument and broken PretrainedConfig type checking are different v5 regressions." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "Both concern attention behavior, but GPT2 attention scaling being ignored is not the same bug as Gemma4 building causal masks for bidirectional mode." + "reason": "GPT2 attention scaling being ignored and Gemma4 bidirectional masking are different attention semantics bugs." }, { "left": "issue:44466", "right": "issue:46002", "accept": false, - "reason": "Serialization inconsistency for lm_head.weight and NaNs in classification heads are different symptoms and code paths." + "reason": "lm_head serialization inconsistency is unrelated to NaNs in classification heads." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and produce different failures." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and likely different code paths." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Missing Qwen3Next projections and Qwen3.5 from_pretrained errors are different model-definition/loading problems." + "reason": "Missing Qwen3Next projections and Qwen3.5 MoE from_pretrained errors are related only by model family, not the same bug." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "GLM-5 GenerationConfig invalid is unrelated to Qwen3.5 flash-attention generation failures." + "reason": "Invalid GenerationConfig and Qwen3.5 flash-attention generation failure are not the same underlying issue." }, { "left": "issue:43450", "right": "issue:43986", "accept": false, - "reason": "A batched video-processor shape bug is not the same as crashing without torchvision in AutoProcessor." + "reason": "Batched video processor shape errors and AutoProcessor dependency crashes are different video-processing problems." }, { "left": "issue:45071", "right": "issue:47059", "accept": false, - "reason": "PretrainedConfig type checking and a CLI completion attribute error from huggingface_hub are unrelated." + "reason": "PretrainedConfig typing regressions and hf hub CLI completion errors are unrelated." }, { "left": "issue:45341", "right": "issue:46328", "accept": false, - "reason": "A testing-utils bug and a SequenceFeatureExtractor.pad inefficiency are not the same defect." + "reason": "A testing_utils bug and SequenceFeatureExtractor.pad inefficiency do not describe the same defect." }, { "left": "issue:45071", "right": "issue:46512", "accept": false, - "reason": "Type-checking breakage and stale config.dtype after from_config are different configuration issues." + "reason": "Type-checking breakage in v5.4.0 and stale config.dtype after from_config are different config issues." }, { "left": "issue:44829", "right": "issue:45216", "accept": false, - "reason": "FlashAttention-3 training collapse and Qwen3.5 save_pretrained regression are unrelated bugs." + "reason": "FlashAttention-3 training collapse and Qwen3.5 checkpoint serialization are separate regressions." }, { "left": "issue:45341", "right": "issue:45561", "accept": false, - "reason": "Both touch testing utilities, but one is a generic bug and the other is an xdist file-race; not the same underlying issue." + "reason": "Both mention testing utils, but one is a small utility bug and the other is an xdist race on captured_info.txt." }, { "left": "issue:45632", "right": "issue:46981", "accept": false, - "reason": "trust_remote_code cache collisions and repeated 429 retries are unrelated loader/network issues." + "reason": "Remote-code cache path collisions and repeated 429 retries are unrelated filesystem vs networking issues." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES None handling and a Mistral regex patch AttributeError are different tokenizer-path failures." + "reason": "TOKENIZER_MAPPING_NAMES None handling and the Mistral regex patch crash are distinct tokenizer-loading failures." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "A sam3_video/sam3_tracker warning is unrelated to a KeyError when loading MobileLLM-125M." + "reason": "SAM3 model-type warning and MobileLLM key error are different model-loading problems." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A 3d_parallel.py runtime failure is unrelated to the testing_utils bug." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but for different symbols and different modules." + "reason": "Both are import errors, but they concern different missing names and likely different export regressions." }, { "left": "issue:43122", "right": "issue:46298", "accept": false, - "reason": "Tokenizer behavior changing across versions and a noisy logging message are not the same bug." + "reason": "Tokenizer behavior regression and noisy logging are unrelated." } ] }, @@ -16535,8 +16503,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -16618,20 +16586,20 @@ "issue:43012|issue:43577", "issue:43874|issue:44261" ], - "prepared_review_unit_hash": "856b5528d34257dcc616452d2087086a5fca48d66e14f6fd1deccf481b1206a1", + "prepared_review_unit_hash": "a6006c2a0cbc7e0b6d462a0b72278172a8a344b5bcc89fc12b82576661dee15a", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35360, + "serialized_chars": 35359, "estimated_input_tokens": 8840, "estimated_eval_tokens": 17936 }, @@ -16640,174 +16608,174 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated Transformers issues that only share broad regression/import/model-loading themes. None of the soft pairs looks like the same underlying bug or change, so all soft edges should be rejected.", - "confidence": 0.88, - "canonical_issue_reason": "issue:45278 is the broadest representative report in the set: a general post-upgrade import-regression complaint that matches the cluster\u2019s dominant theme better than the model-specific or one-off bugs.", + "summary": "The cluster is mostly heterogeneous: the soft pairs share broad Transformers themes (tokenizers, imports, model loading, FlashAttention, generation, etc.) but do not look like the same concrete bug/change. I found no convincing duplicate pairs to merge.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: it describes multiple import regressions after a version upgrade, making it the closest thing to a cluster representative among otherwise mostly unrelated issue reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best cluster anchor because it is the most general and user-facing regression report, making it the clearest representative issue even though it is not a duplicate of the others.", + "best_issue_reason": "issue:45278 is the best representative issue because it summarizes a wide, user-visible regression class and is more likely than the others to subsume neighboring import-related complaints.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45362", "right": "issue:46328", "accept": false, - "reason": "Different problems: Qwen3.5 chat crashing vs a SequenceFeatureExtractor.pad performance quirk." + "reason": "Different subsystems and failures: Qwen3.5 chat inference crash vs SequenceFeatureExtractor padding performance." }, { "left": "issue:42222", "right": "issue:46178", "accept": false, - "reason": "Different code paths: vitpose model breakage vs a backend mapping error for tensorflow_text." + "reason": "VitPose model breakage is unrelated to the tensorflow_text backend mapping error." }, { "left": "issue:39401", "right": "issue:46209", "accept": false, - "reason": "Tokenizer offset_mapping bug vs FP8 inference failure on B200; not the same bug." + "reason": "Tokenizer offset mapping bug and FP8/B200 inference failure are unrelated code paths." }, { "left": "issue:44877", "right": "issue:46231", "accept": false, - "reason": "Loading granite_speech config vs documentation mismatch for set_training(); unrelated." + "reason": "Granite speech config loading and a documentation parameter-name mismatch are not the same bug." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API guidance vs MLA q_a_layernorm precision issue; not the same defect." + "reason": "TF32 API guidance and Minimax MLA layernorm epsilon behavior are unrelated." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "A save_pretrained shard-size question vs a Qwen3.5 MoE from_pretrained error are different issues." + "reason": "Save_pretrained sharding limit question vs Qwen3.5 MoE from_pretrained failure; no shared bug." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "Both are model-loading errors, but for different models and different failure modes." + "reason": "Invalid GenerationConfig and unrecognized model/config loading are different failures." }, { "left": "issue:44261", "right": "issue:46762", "accept": false, - "reason": "Both mention numerical behavior, but they concern different models/components and different symptoms." + "reason": "Both mention numerical issues, but one is a layernorm-epsilon problem and the other is an attention-vs-reference mismatch; too different to merge." }, { "left": "issue:42371", "right": "issue:46617", "accept": false, - "reason": "TF32 configuration discussion vs cache FileNotFoundError for trust_remote_code; unrelated." + "reason": "TF32 settings and trust_remote_code cache FileNotFoundError are unrelated." }, { "left": "issue:44933", "right": "issue:45278", "accept": false, - "reason": "Specific missing import from image_utils vs a broad version-upgrade import regression; too different to merge." + "reason": "A specific missing import from image_utils is not the same concrete bug as a broad post-upgrade import-regression report." }, { "left": "issue:38617", "right": "issue:44220", "accept": false, - "reason": "Missing layer_type_validation import vs _torch_extract_fbank_features issue; distinct failures." + "reason": "Missing layer_type_validation import and _torch_extract_fbank_features are unrelated." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Different Qwen variants and different symptoms: missing projections vs from_pretrained error." + "reason": "Different Qwen model families and different failures: missing projections vs from_pretrained error." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "SGD arguments ignored vs optimizer.step() called twice under deepspeed; different optimizer bugs." + "reason": "SGD arguments ignored is a different optimizer bug from deepspeed causing double optimizer steps." }, { "left": "issue:43066", "right": "issue:43122", "accept": false, - "reason": "Wrong tokenizer decoder type vs changed tokenization output across versions; related area but not the same underlying bug." + "reason": "Both involve tokenizer behavior in v5, but one is an internal decoder-type issue and the other is a broader tokenization difference; not clearly the same concrete bug." }, { "left": "issue:43208", "right": "issue:46894", "accept": false, - "reason": "xLSTM training blockers vs CohereASR loss double-shift; completely different model bugs." + "reason": "xLSTM training blockers and CohereASR double-shift loss bug are unrelated." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "torch.split return-value handling vs RoPE implementation issue; different internal logic." + "reason": "torch.split return handling in GLM Moe indexing is unrelated to the GLM-5 RoPE implementation discussion." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removed tie_embeddings_and_encoder_decoder config flag vs unexpected position-id keys; different breakages." + "reason": "Removed tie_embeddings_and_encoder_decoder and unexpected position-id keys are different regressions." }, { "left": "issue:44077", "right": "issue:46981", "accept": false, - "reason": "patchtsmixer post_init API restriction vs repeated HTTP 429 retries in AutoProcessor; unrelated." + "reason": "PatchTSMixer post_init policy and AutoProcessor 429 retry behavior are unrelated." }, { "left": "issue:43673", "right": "issue:45341", "accept": false, - "reason": "Missing cache in chunked_prefill vs a testing_utils bug; not the same problem." + "reason": "GenerationMixin cache during chunked_prefill and a testing_utils bug are unrelated." }, { "left": "issue:44483", "right": "issue:45464", "accept": false, - "reason": "Chat/completions request rejection vs streaming inference failure; same broad area but different concrete bugs." + "reason": "Both concern chat/completions, but one is a general endpoint acceptance issue and the other is a Qwen3.5 streaming inference failure; not the same bug." }, { "left": "issue:43931", "right": "issue:44410", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch vs missing projections in qwen3next; unrelated model-loading issues." + "reason": "Qwen3-VL weight-shape mismatch and qwen3next missing projections are distinct model-loading/model-architecture issues." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "AutoImageProcessor import failure vs PretrainedConfig type-checking break; distinct regression types." + "reason": "AutoImageProcessor import failure and PreTrainedConfig type-checking breakage are unrelated." }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "Both are Flash Attention 2-related, but for different models and different support gaps." + "reason": "Both mention Flash Attention 2, but they affect different models and likely different support gaps." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "Save shard-size question vs Qwen3.5 MoE from_pretrained error; no shared underlying bug." + "reason": "Save sharding default question vs Qwen3.5 MoE from_pretrained error are unrelated." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "FlashAttention failure in Ministral-3 vs degenerate training with flash_attention_3; separate model-specific issues." + "reason": "Ministral FlashAttention failure and flash_attention_3 degenerate training are different concrete issues." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "PyTorch warning during bf16 compile vs dtype fields staying float32 in Blip2; related precision space but different bugs." + "reason": "A compile-time bf16 warning and Blip2 dtype not propagating to qformer are different bugs." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Missing image-patch helper in GLM46V processor vs MLA layernorm epsilon issue; different multimodal failures." + "reason": "Missing image-patch helper in GLM46V and Minimax layernorm epsilon handling are unrelated." } ] }, @@ -16826,8 +16794,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -16905,234 +16873,234 @@ "issue:43986|issue:44484", "issue:46655|issue:46664", "issue:46298|issue:46752", + "issue:42890|issue:46298", "issue:38617|issue:44351", "issue:44062|issue:44117", - "issue:42890|issue:46298", "issue:43097|issue:43873", "issue:43012|issue:46821", "issue:44117|issue:46701", "issue:46172|issue:46710" ], - "prepared_review_unit_hash": "c47dfad96f2baacb63919a44f606e402167850b22595a58fc4ffb1407af07d18", + "prepared_review_unit_hash": "de9ec07b50adcb2d60717b73d0d40bf15ece47cd3a2e543f7e9e47e4cb730e8f", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 36512, - "estimated_input_tokens": 9128, - "estimated_eval_tokens": 18512 + "serialized_chars": 36513, + "estimated_input_tokens": 9129, + "estimated_eval_tokens": 18514 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The soft pairs are mostly broad-similarity false positives across different modules, models, or failure modes; none are safe duplicates. If a representative issue is needed, issue:43873 is the strongest general target.", - "confidence": 0.94, - "canonical_issue_reason": "issue:43873 is the broadest and most active open report in this set, with repeated discussion/inbound references around quantization/offloading behavior, so it is the best representative issue.", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers bug reports. None of the soft-pair candidates look like the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.76, + "canonical_issue_reason": "Issue #43873 is the most developed and actively discussed report in the set, so it is the best single anchor if one issue must represent the cluster; however, the cluster itself is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 has the widest apparent impact and the most ongoing attention, making it the strongest standalone issue to anchor this cluster.", + "best_issue_reason": "Issue #43873 has the richest context, is still open, and describes a concrete actionable regression, making it the strongest overall issue in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44263", "right": "issue:46172", "accept": false, - "reason": "Different model families and code paths: GlmMoeDsaIndexer split semantics vs DeepSeek-v4 dtype/transpose." + "reason": "Different bugs in DSA/xLSTM-style indexing vs DeepSeek-V4 tensor handling; no shared code path." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "DSA ReLU omission is unrelated to DeepSeek-v4 tensor transpose behavior." + "reason": "Both mention DSA/indexing, but one is a missing ReLU in the indexer and the other is a dtype/transpose bug." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Both mention Qwen3-VL, but one is a loading shape mismatch and the other a broader Qwen3-VL-Moe/transformers 5.0 issue; not the same defect." + "reason": "Same broader Qwen3-VL area, but one is a model weight-shape mismatch and the other is a Qwen3-VL-Moe bug." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "API removal of tie_embeddings_and_encoder_decoder is unrelated to gguf_file from_pretrained failures." + "reason": "One is a removed config API in v5, the other is gguf loading failing in from_pretrained; different failure modes." }, { "left": "issue:43986", "right": "issue:45341", "accept": false, - "reason": "Video AutoProcessor crash without torchvision and a testing_utils bug are different code paths." + "reason": "AutoProcessor/torchvision import crash vs a testing_utils bug; unrelated code paths." }, { "left": "issue:45003", "right": "issue:45341", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to the testing_utils issue." + "reason": "sys.modules access in modeling_utils vs a testing helper bug are separate issues." }, { "left": "issue:42886", "right": "issue:46981", "accept": false, - "reason": "Offline cache loading failure and repeated HTTP 429 retries for HF_ENDPOINT are distinct behaviors." + "reason": "Both involve Hub/network behavior, but offline cache loading and repeated 429 retries are distinct problems." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "MLA norm-epsilon precision drift is unrelated to GLM-5 RoPE implementation." + "reason": "MLA rms_norm_eps/config precision issue vs GLM-5 RoPE implementation; not the same root cause." }, { "left": "issue:45753", "right": "issue:46369", "accept": false, - "reason": "Qwen3_5 infinite-loop behavior and SAM memory overflow are different bugs in different models." + "reason": "Qwen3_5 infinite-loop bug on a specific image vs SAM3 memory overflow; different models and symptoms." }, { "left": "issue:44117", "right": "issue:44987", "accept": false, - "reason": "Tokenizer mapping None handling and physical-intelligence/fast loading failure are different root causes." + "reason": "Tokenizer mapping None handling vs a specific transformers>=5.1 loading failure; separate loading bugs." }, { "left": "issue:43531", "right": "issue:45341", "accept": false, - "reason": "Qwen3-MoE sliding_window bug is unrelated to the testing_utils report." + "reason": "Sliding_window issue in Qwen3-MoE vs a testing utility bug; not the same bug." }, { "left": "issue:43097", "right": "issue:45588", "accept": false, - "reason": "Removed API surface and flash_attention s_aux=None crash are different subsystems and symptoms." + "reason": "v5 API removal vs flash_attention crashing on sink-less models; distinct code paths." }, { "left": "issue:43673", "right": "issue:46123", "accept": false, - "reason": "Generation cache/chunked_prefill issue is unrelated to the mask pipeline final-batch flag bug." + "reason": "GenerationMixin cache missing during chunked_prefill vs partial-batch dropping in MaskGenerationPipeline; different defects." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Image-patch counting and torch.split return-value handling are separate defects." + "reason": "Missing get_number_of_image_patches vs torch.split return values; separate attribute and tensor-shape bugs." }, { "left": "issue:45440", "right": "issue:46172", "accept": false, - "reason": "Both are DeepSeek-family reports, but remote-implementation divergence and a specific DeepSeek-v4 tensor-op issue are not the same bug." + "reason": "Native DeepseekV3MoE divergence vs a DeepSeek-V4 tensor/transposition issue; different implementations." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Missing attention projections and missing vision config fields are unrelated model-definition issues." + "reason": "Missing attn projections in qwen3next vs a missing config field silently dropped; different layers and failure modes." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Video processor torchvision dependency crash and mistral_common ReasoningEffort import error are distinct failures." + "reason": "Video-model loading crash without torchvision vs Gemma 4 processor import error from mistral_common; unrelated dependencies." }, { "left": "issue:44038", "right": "issue:44410", "accept": false, - "reason": "Qwen3-VL-Moe bug and Qwen3Next layer-0 projection omission are different concrete problems." + "reason": "Both are Qwen3-family, but one is a Qwen3-VL-Moe issue and the other is missing projections in qwen3next." }, { "left": "issue:43097", "right": "issue:46050", "accept": false, - "reason": "API removal is unrelated to quantized save/load corruption after sinq." + "reason": "Removed API in v5 vs quantized model save/load corruption; not the same underlying bug." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are ImportErrors, but for different missing symbols from different modules." + "reason": "Both are import errors, but for different modules and symbols." }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "FSDP2 optimizer-state corruption and Liger Kernel not being applied are unrelated training bugs." + "reason": "FSDP2/PEFT optimizer-state bug vs Liger Kernel not applied with model_init; separate training-path issues." }, { "left": "issue:43208", "right": "issue:46897", "accept": false, - "reason": "Different training bugs in different models; only loosely similar at a high level." + "reason": "xLSTM training blockers vs Florence2 double-shift loss bug; different models and mechanics." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "Processor import crash without torchvision is unrelated to save_pretrained shard-size behavior." + "reason": "AutoProcessor import crash vs a save_pretrained shard-size question; unrelated." }, { "left": "issue:46655", "right": "issue:46664", "accept": false, - "reason": "Same SmolVLM video processor area, but padding calculation and num_frames semantics are separate defects." + "reason": "Both are SmolVLM video-processor issues, but padding math and num_frames semantics are different bugs." }, { "left": "issue:46298", "right": "issue:46752", "accept": false, - "reason": "Noisy logging and apply_chat_template empty-list IndexError are unrelated." + "reason": "Noisy logging message vs apply_chat_template crashing on an empty list; unrelated." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "SAM HQ test flakiness vs logging noise; not the same bug." }, { "left": "issue:38617", "right": "issue:44351", "accept": false, - "reason": "Different removed imports from different modules." + "reason": "Both are import errors, but they involve different public APIs and packages." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "AddedToken special-argument collision and tokenizer-mapping fallback failure are different bugs." - }, - { - "left": "issue:42890", - "right": "issue:46298", - "accept": false, - "reason": "Integration test seed flakiness and a noisy logging message are unrelated." + "reason": "AddedToken special-kwarg conflict vs tokenizer-mapping None handling; separate tokenizer initialization bugs." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "Removed API surface and quantization/offloading regression are different problems." + "reason": "v5 API removal vs quantization/offloading behavior; distinct code paths and symptoms." }, { "left": "issue:43012", "right": "issue:46821", "accept": false, - "reason": "PyTorch compile warning and TimesFM window_size AttributeError are unrelated." + "reason": "PyTorch warning during bfloat16 compilation vs TimesFM window_size AttributeError; unrelated." }, { "left": "issue:44117", "right": "issue:46701", "accept": false, - "reason": "Tokenizer mapping fallback bug and image-text-to-text pipeline failure are different code paths." + "reason": "Tokenizer mapping fallback bug vs image-text-to-text pipeline failure for DiffusionGemmaGenerationOutput." }, { "left": "issue:46172", "right": "issue:46710", "accept": false, - "reason": "DeepSeek-v4 tensor-op issue and DeepSeek-R1 accuracy regression after upgrade are not the same concrete bug." + "reason": "DeepSeek-V4 tensor-construction bug vs an accuracy regression after upgrade; not the same root cause." } ] }, @@ -17151,8 +17119,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -17227,8 +17195,8 @@ "issue:45584|issue:46365", "issue:44380|issue:46132", "issue:44162|issue:46172", - "issue:44077|issue:46365", "issue:42890|issue:47059", + "issue:44077|issue:46365", "issue:44222|issue:45478", "issue:36683|issue:43582", "issue:44380|issue:45479", @@ -17239,20 +17207,20 @@ "issue:44483|issue:44514", "issue:38617|issue:43502" ], - "prepared_review_unit_hash": "f6312a24421ab3b695bcd4ea0a4051c86a4672914b7aeffb1a6966d3de909286", + "prepared_review_unit_hash": "ac49c987367b0b10b22237e49e9d54c85758325c435c94e2119fe3114bf33f4b", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37225, + "serialized_chars": 37226, "estimated_input_tokens": 9307, "estimated_eval_tokens": 18870 }, @@ -17261,204 +17229,204 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft edge looks like a true duplicate or single mergeable fix; the candidates span unrelated bugs across loading, saving, training, compilation, docs, and environment issues. No PRs were provided. If a representative standalone issue is needed, #43502 is the most broadly actionable user-facing regression among the set.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "The soft pairs are mostly superficial similarity matches across unrelated bugs (imports, warnings, saving, backend behavior, and model-specific failures). None of the candidate pairs look like the same underlying issue or a mergeable PR duplicate.", + "confidence": 0.83, + "canonical_issue_reason": "issue:44380 is the most connected node in the soft-pair graph and is a well-scoped concrete runtime bug, making it the best representative issue among a very heterogeneous set.", "canonical_pr_reason": null, - "best_issue_reason": "#43502 is the strongest standalone representative: it is a concrete, broadly relevant loading bug with clear repro impact, while the other issues are narrower, model-specific, or environment-specific.", + "best_issue_reason": "issue:44380 has the highest apparent centrality and describes a specific code-path regression, so it is the most suitable global representative even though the cluster is not truly duplicate-cohesive.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "Different bugs: one is unintended network access with local_files_only, the other is a missing embed_tokens.weight during UMT5Encoder loading." + "reason": "One is a local-files-only network/request bug; the other is a missing model weight during from_pretrained. Different failure modes and code paths." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "Different code paths: AutoProcessor kwargs forwarding vs auto_docstring crashing on future annotations." + "reason": "AutoProcessor kwargs forwarding and auto-docstring parsing with future annotations are unrelated subsystems." }, { "left": "issue:44936", "right": "issue:46365", "accept": false, - "reason": "Different failures: trainer.evaluate after train vs importing Trainer breaking until datasets is removed." + "reason": "Trainer.evaluate() after train() is not the same as an import failure caused by datasets shadowing Trainer imports." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "Different problems: GPT2 attention scaling ignored in SDPA/FlashAttention vs cache lazy init causing recompilation with chunked prefill." + "reason": "Attention scaling being ignored under SDPA/FlashAttention is unrelated to cache lazy initialization and chunked prefill recompilation." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Unrelated: Python 3.14 import failure vs a backend mapping error for tensorflow_text." + "reason": "Python 3.14 import compatibility and an invalid backend mapping for tensorflow_text are different root causes." }, { "left": "issue:43986", "right": "issue:44016", "accept": false, - "reason": "Different areas entirely: video-model AutoProcessor import crash vs a syntax error in a notebook." + "reason": "Missing torchvision import handling is unrelated to a notebook syntax error." }, { "left": "issue:42907", "right": "issue:45753", "accept": false, - "reason": "Different model behaviors: save failure for dequantized Ministrals/Devstrals vs Qwen3_5 infinite loop on a specific image." + "reason": "Saving dequantized models and an image-triggered infinite generation loop are different bugs." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Both involve saving, but not the same bug: model dequantization save failure is not the same as saving Pipeline objects." + "reason": "Pipeline save_pretrained failure and dequantized model save failure are both save-related but not the same underlying bug." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "Different symptoms and paths: a precision warning during compile vs an actual torch.compile failure with SDPA." + "reason": "A warning about bfloat16/float32 precision is not the same as a torch.compile SDPA failure on a specific model." }, { "left": "issue:45712", "right": "issue:46298", "accept": false, - "reason": "Different scope: cleanup of leftover dummy classes vs a noisy logging message." + "reason": "Dead dummy classes in a repo audit are unrelated to noisy logging output." }, { "left": "issue:43012", "right": "issue:44368", "accept": false, - "reason": "Both are warnings, but about different configurations: bfloat16 compile warning vs tie_word_embeddings warning." + "reason": "Both mention warnings around embeddings, but one is a PyTorch precision warning and the other is a model/config warning about tie_word_embeddings." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "Different save/load issues: missing embed_tokens.weight in from_pretrained vs max_shard_size behavior in save_pretrained." + "reason": "Missing embed_tokens.weight during load is unrelated to the default max_shard_size in save_pretrained." }, { "left": "issue:43208", "right": "issue:43550", "accept": false, - "reason": "Different model families and failures: xLSTM training bugs vs Bamba compile/SDPA failure." + "reason": "xLSTM training blockers are model-specific and do not match a Bamba compile/SDPA regression." }, { "left": "issue:43825", "right": "issue:44655", "accept": false, - "reason": "Different APIs: pipeline task-support messaging vs inability to save Pipeline objects." + "reason": "A misleading pipeline error message and inability to save pipeline objects are different issues." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Both mention attention backends, but they are different model/backend bugs and not one fixable code-path issue." + "reason": "A model explicitly not supporting Flash Attention 2.0 is not the same as GPT2 ignoring attention scaling under backends." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "Different problems: FlashAttention support gating vs a Python version compatibility issue." + "reason": "A model/backend capability issue is unrelated to Python version compatibility during import." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Not enough evidence of the same bug; wrong tokenizer decoder type is unrelated to an isolated reproduction of another issue." + "reason": "Wrong tokenizer decoder type in v5 and an isolated reproduction of another issue do not describe the same bug." }, { "left": "issue:43097", "right": "issue:46298", "accept": false, - "reason": "Different changes: a removed API without notice vs a noisy logging message." + "reason": "A removed API with no indication is unrelated to a noisy logging message." }, { "left": "issue:45584", "right": "issue:46365", "accept": false, - "reason": "Different subsystems: Whisper generation edge case vs Trainer import being affected by datasets." + "reason": "Whisper empty-transcription generation failure and a Trainer import conflict are unrelated." }, { "left": "issue:44380", "right": "issue:46132", "accept": false, - "reason": "Both touch attention internals, but one is backend scaling behavior and the other is register() changing function behavior." + "reason": "Attention scaling behavior under SDPA/FlashAttention is not the same as register() mutating a function's behavior." }, { "left": "issue:44162", "right": "issue:46172", "accept": false, - "reason": "Different model-specific failures: ESM2 broken vs a DeepSeek-V4 dtype/transpose issue." + "reason": "ESM2 breakage and a DeepSeek tensor transpose/dtype bug are model-specific and distinct." }, { - "left": "issue:44077", - "right": "issue:46365", + "left": "issue:42890", + "right": "issue:47059", "accept": false, - "reason": "Different bugs: patchtsmixer post_init validation vs Trainer import failure." + "reason": "Flaky test seeding and a cli incompatibility with huggingface_hub 1.22.0 are unrelated." }, { - "left": "issue:42890", - "right": "issue:47059", + "left": "issue:44077", + "right": "issue:46365", "accept": false, - "reason": "Unrelated: flaky integration tests from missing set_seed vs CLI tests breaking with huggingface_hub 1.22.0." + "reason": "patchtsmixer post_init validation and an import conflict with datasets have no shared underlying code path." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "Related to MoE models, but not the same concrete bug: FP8 save_pretrained vs from_pretrained error on Qwen3.5 MoE." + "reason": "FP8 save_pretrained behavior and a Qwen3.5 MoE from_pretrained error are different model lifecycle bugs." }, { "left": "issue:36683", "right": "issue:43582", "accept": false, - "reason": "Completely different domains: Gemma3Config missing vocab_size vs an Apple Silicon TypeError in caching_allocator_warmup." + "reason": "Gemma3Config missing vocab_size and an Apple Silicon TypeError in caching_allocator_warmup are unrelated." }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Different bugs: attention scaling ignored vs degenerate zero loss in sequence classification." + "reason": "Attention scaling being ignored is unrelated to a degenerate zero-loss classification setup." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are ImportError reports, but for different symbols and unrelated modules." + "reason": "ImportError for is_offline_mode and missing Qwen2_5_VLForConditionalGeneration are different missing-symbol problems." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "Different training/distributed optimizer-state bug vs GPT2 attention scaling behavior." + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT2 attention scaling under SDPA/FlashAttention." }, { "left": "issue:42175", "right": "issue:46365", "accept": false, - "reason": "Different dependency/import problems: missing TensorFlow backend in extras install vs Trainer import conflicting with datasets." + "reason": "Backend packaging for tensorflow versus Trainer import failure due to datasets are both dependency issues but not the same bug." }, { "left": "issue:42490", "right": "issue:43012", "accept": false, - "reason": "Different symptoms: local model path slash warning vs a compile-time precision warning." + "reason": "A trailing-slash local model warning and a bfloat16 compile precision warning are different concerns." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "Different request surfaces: chat/completions rejection vs batched apply_chat_template crash." + "reason": "Chat/completions request rejection and batched chat template crashes are different endpoints and failure modes." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "Different issues: missing configuration_utils symbol import vs unintended API requests with local_files_only." + "reason": "A missing configuration_utils import and unexpected API requests despite local_files_only=True are unrelated." } ] }, @@ -17477,8 +17445,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -17565,20 +17533,20 @@ "issue:44908|issue:46851", "issue:29870|issue:33453" ], - "prepared_review_unit_hash": "9c4bdb0fda93cf4dafac020085e1735f244a0698ac8d6d2ec21f6dd271742ea9", + "prepared_review_unit_hash": "dfb30089cd59dd2952e9369303327e181d5715c0d1501fed0ab665c6294d1131", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 36879, + "serialized_chars": 36877, "estimated_input_tokens": 9220, "estimated_eval_tokens": 18696 }, @@ -17587,204 +17555,204 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is thematically related (regressions/import/load/save issues) but the soft-edge pairs mostly describe different concrete bugs and code paths, so none should be merged as duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "If one issue must anchor the cluster, issue:45278 is the broadest umbrella report: a version-upgrade wave of import/load failures. The other items are narrower model-specific or API-specific regressions.", + "summary": "These items are mostly unrelated regressions, import/version incompatibilities, and model-specific bugs. I don\u2019t see a true duplicate cluster here; the soft similarities are mostly broad subsystem overlap rather than the same underlying issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:36683 is the most active concrete bug report here and a reasonable anchor for the set, with strong discussion activity and a clearly described model/config regression.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best representative because it captures the widest user-visible regression surface in the set, while the rest are more specialized instances of failure.", + "best_issue_reason": "issue:36683 is the best single issue to represent the cluster because it is open, highly discussed, and more concrete than the surrounding environment/import-only reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43618", "right": "issue:44380", "accept": false, - "reason": "Both are regressions, but one is CLIP attentions output assignment and the other is GPT2 attention scaling under SDPA/FlashAttention; different code paths." + "reason": "Both involve attention behavior, but one is missing CLIP attentions while the other ignores GPT-2 attention scaling config; different code paths and fixes." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Smollm3 RoPE-layer count bug and GLM-5 RoPE implementation discussion are related only at the model-family level, not the same defect." + "reason": "Both mention RoPE, but they describe different model implementations and different symptoms; not the same underlying bug." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "xLSTM training blockers and BeitImageProcessorFast label reduction are unrelated bugs in different components." + "reason": "Completely unrelated: xLSTM training bugs versus a BeitImageProcessorFast label-reduction issue." }, { "left": "issue:43618", "right": "issue:43761", "accept": false, - "reason": "Both touch CLIP regressions, but one is about attentions not being assigned and the other about hidden_states being dropped; not the same failure." + "reason": "Both are CLIP v5 regressions, but one loses attentions and the other returns hidden_states=None; related theme, not clearly the same bug." }, { "left": "issue:45276", "right": "issue:46077", "accept": false, - "reason": "Gemma4 resize_token_embeddings propagation and Gemma4 bidirectional attention masking are separate bugs in different parts of the model." + "reason": "Different Gemma4 problems: token embedding resize propagation versus bidirectional-attention mask construction." }, { "left": "issue:36010", "right": "issue:42757", "accept": false, - "reason": "Different dependency breakages: GenerationMixin import from transformers.generation vs is_offline_mode import from huggingface_hub." + "reason": "Both are import errors, but they come from different modules and different missing symbols; separate compatibility issues." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Qwen3VL MoE loading regression and sorted-state_dict load failure are both load-time issues, but the symptoms and root causes are different." + "reason": "Both are model-loading failures, but for different failure modes and likely different root causes." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Quantized-model saving failure and Pipeline.save_pretrained failure are distinct save-pretrained problems affecting different object types." + "reason": "Both involve saving, but quantized-model save failure and Pipeline.save_pretrained failure are different problems." }, { "left": "issue:44038", "right": "issue:44230", "accept": false, - "reason": "One reports a general Qwen3-VL-Moe bug, the other asks for FP8 inference support; not the same concrete bug." + "reason": "Same family of Qwen3-VL-Moe topics, but one is a transformers regression and the other is fp8 inference support; not the same change." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "A chat-completions API request rejection and an apply_chat_template crash are different request-processing paths." + "reason": "Different APIs and symptoms: chat/completions request handling versus apply_chat_template crashing on tool-call assistant messages." }, { "left": "issue:44625", "right": "issue:46365", "accept": false, - "reason": "Qwen3.5 num_labels propagation and Trainer import failure with datasets are unrelated." + "reason": "Unrelated: Qwen3.5 num_labels propagation versus a Trainer import conflict with datasets." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Nemotron Flash Attention support and GLM-5 GenerationConfig validation are separate model-specific issues." + "reason": "Different models and failures: Flash Attention 2 support gating versus an invalid GenerationConfig issue." }, { "left": "issue:44355", "right": "issue:46877", "accept": false, - "reason": "Compiled Python file loading errors and auto_docstring stdout noise at import time are unrelated." + "reason": "One is about compiled Python files failing, the other about import-time docstring lint output; unrelated." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Different missing-attribute bugs on different objects/models: Gemma3Config.vocab_size vs Sam3VisionEncoderOutput.fpn_position_embeddings." + "reason": "Different missing attributes on different model outputs; same broad pattern, but not the same bug." }, { "left": "issue:43611", "right": "issue:44655", "accept": false, - "reason": "Base_model_prefix loading regression and Pipeline.save_pretrained failure are both load/save related, but not the same bug." + "reason": "Base-model-prefix loading regression and Pipeline.save_pretrained failure are separate issues." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "Qwen3.5 num_labels propagation and Gemma4 attention-mask construction are unrelated model bugs." + "reason": "Different Gemma4/Qwen3.5 bugs with no shared code-path indicated." }, { "left": "issue:44938", "right": "issue:47059", "accept": false, - "reason": "Python 3.14 loading failure and cli completion API breakage from huggingface_hub are different compatibility issues." + "reason": "Python 3.14 import failure and a huggingface_hub CLI compatibility failure are different environment breakages." }, { "left": "issue:44655", "right": "issue:46050", "accept": false, - "reason": "Pipeline serialization and sinq quantized model round-trip are distinct save/load failures." + "reason": "Both are save/load related, but one is Pipeline serialization and the other is sinq quantization round-trip failure." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Both mention Python-version compatibility, but they affect different releases and failure modes." + "reason": "Both are version-compatibility problems, but they target different Python releases and likely different causes." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Sorted-state_dict loading and Qwen3.5 checkpoint corruption after save_pretrained are different persistence bugs." + "reason": "One is about sorted state_dict loading order, the other about Qwen3.5 save_pretrained checkpoint correctness." }, { "left": "issue:36683", "right": "issue:44077", "accept": false, - "reason": "Gemma3Config missing vocab_size and xLSTM training blockers are unrelated." + "reason": "Unrelated: missing Gemma3Config attribute versus a patchtsmixer post_init policy change." }, { "left": "issue:46077", "right": "issue:46082", "accept": false, - "reason": "Gemma4 attention-mask behavior and LlamaConfig head_dim validation are separate configuration semantics." + "reason": "Different model configuration validations: attention-mask behavior versus head_dim constraints." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "CLIP attentions regression and Qwen3.5 checkpoint save regression do not share the same code path." + "reason": "CLIP output regression and Qwen3.5 save_pretrained regression are unrelated." }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Both are missing-attribute errors, but they involve different configs and different models." + "reason": "Different missing config attributes on different models; not the same bug." }, { "left": "issue:36683", "right": "issue:44625", "accept": false, - "reason": "Gemma3Config.vocab_size missing and Qwen3.5 num_labels propagation are different configuration issues." + "reason": "Gemma3Config vocab_size and Qwen3.5 num_labels are unrelated configuration issues." }, { "left": "issue:29870", "right": "issue:36296", "accept": false, - "reason": "BART generate decoder_attention_mask bug and tensor-parallel training bug are unrelated failures." + "reason": "Bart generate() UnboundLocalError and tensor-parallel training bug are different failures." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "Wrong optimizer states on non-zero ranks and a cache-strategy typo are not the same issue." + "reason": "Training optimizer-state corruption and a typo in cache strategies are unrelated." }, { "left": "issue:43010", "right": "issue:45278", "accept": false, - "reason": "torch.no_grad decoration for cache update methods and broad import errors after upgrade are unrelated." + "reason": "A no_grad decoration issue in cache updates is not the same as broad import failures after upgrading." }, { "left": "issue:44075", "right": "issue:46365", "accept": false, - "reason": "Unused SGD args and Trainer import failure due to datasets are different problems." + "reason": "SGD argument handling and Trainer import conflicts are unrelated." }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "UMT5Encoder missing embed_tokens.weight and Qwen3.5 MoE from_pretrained error are separate model-loading bugs." + "reason": "Both are from_pretrained/load issues, but they involve different models and distinct missing-weight behaviors." }, { "left": "issue:44908", "right": "issue:46851", "accept": false, - "reason": "Scheduler kwargs handling and variable-length position_ids in Qwen3.5 are unrelated." + "reason": "Scheduler kwargs are ignored versus variable-length Qwen3.5 position_ids handling; unrelated." }, { "left": "issue:29870", "right": "issue:33453", "accept": false, - "reason": "BART generate decoder_attention_mask error and tokenizer loading regression are different bugs." + "reason": "Bart generate() decoder_attention_mask error and tokenizer-loading regression are separate bugs." } ] }, @@ -17803,8 +17771,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -17890,20 +17858,20 @@ "issue:30333|issue:33453", "issue:45125|issue:45478" ], - "prepared_review_unit_hash": "afcfd400748c70f28895ab9512f689fd318e716f730b3a9001163c56cee82645", + "prepared_review_unit_hash": "43f2d506d8f4dd62322e69b9ecb785a3433715c43a989441ec0feeaaa721a3e9", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36696, + "serialized_chars": 36694, "estimated_input_tokens": 9174, "estimated_eval_tokens": 18604 }, @@ -17912,208 +17880,202 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue reports across loading, serialization, tensor-parallelism, perf, and flaky-test regressions. Only one soft edge looks like a likely true duplicate: the Qwen3.5 MoE tensor-parallel load failure pair.", - "confidence": 0.64, - "canonical_issue_reason": "issue:45125 is the clearest root-cause report: it names the missing `_tp_plan` tensor-parallel metadata on Qwen3.5 MoE, which likely underlies the paired `from_pretrained` failure.", + "summary": "Heterogeneous cluster with a loose concentration around v5 loading/regression complaints, but no soft pair is mergeable; the previously overcalled 44393/47000 pair is not accepted.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43723 is the broadest user-facing loading-regression representative in the set and best captures the cluster\u2019s most common theme without overfitting to a niche model family.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45125 is the most actionable and specific issue in the set; it points to the concrete bug rather than just the downstream load error.", + "best_issue_reason": "issue:43723 is the strongest global issue choice because it is generic, high-signal, and closest to the dominant loading/failure pattern across the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44393", "right": "issue:47000", "accept": false, - "reason": "Different model/task failures: 2D bbox hallucination vs RTDetr loss infeasible cost matrix." + "reason": "Different model families and failure modes: Qwen3-VL 2D box hallucination vs RTDetrLoss cost-matrix infeasibility crash." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "Both involve model loading/roundtrip, but they target different models and different failure modes." + "reason": "Both are load/roundtrip-related at a high level, but the concrete bugs differ: sorted state_dict loading vs Gemma pipeline roundtrip on a token character." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Both are import/compile-time problems, but one is compiled Python files and the other is a TorchScript parsing issue." + "reason": "Unrelated import/runtime failures: compiled Python file handling vs a Python 3.13 TorchScript parse/IndentationError." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "Unrelated bugs: Gemma3 config missing `vocab_size` vs FSDP2/PEFT optimizer state corruption." + "reason": "Different subsystems and symptoms: missing Gemma3Config attribute vs distributed optimizer-state corruption under FSDP2/PEFT." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "Both are load/serialization issues, but they concern different Qwen models and different missing fields/errors." + "reason": "Both touch loading, but one is a state_dict sorting issue and the other is a missing vision-config field being dropped." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "Both are test flakiness, but one is missing `set_seed()` and the other is an xdist file race." + "reason": "Both are test-flake reports, but one is missing set_seed() in integration tests and the other is an xdist file race." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "A specific `@torch.no_grad` proposal is not the same bug as a broad claim that v5 breaks too many models." + "reason": "No_grad decoration request and broad v5 breakage complaint are different problem scopes and code paths." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Both mention tensor parallelism, but one is a generic training bug and the other is an OOM while loading a specific model." + "reason": "Tensor-parallel is the only overlap; one is a generic training bug, the other is a model-specific OOM when loading Qwen2-57B-A14B-Instruct." }, { "left": "issue:41669", "right": "issue:44246", "accept": false, - "reason": "Shared theme of import slowness, but one is a concrete import-* cleanup issue and the other is a vague intermittent slow-import report." + "reason": "Both mention import slowness, but there is no evidence they share the same root cause or touched code path." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "Different code paths: missing `attentions` assignment vs invalid optional `post_init` usage." + "reason": "Different fixes: CLIPOutput attentions assignment vs patchtsmixer post_init policy." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "`no_grad` on cache updates is unrelated to a Python-version compatibility issue." + "reason": "Completely different concerns: cache/layer no_grad decoration vs Python-version compatibility." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Both are serialization regressions, but they affect different models and different save/load paths." + "reason": "Generic load/save regression vs Qwen3.5 visual-encoder key regression; not the same concrete bug." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Different model families and symptoms: UMT5 encoder weight loading vs Qwen3.5 MoE from_pretrained failure." + "reason": "Both are from_pretrained failures, but they involve different Qwen3.5 models and different loading-path defects." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "PyTorch missing import error is unrelated to tokenizer-loading regression." + "reason": "Missing PyTorch dependency import error is unrelated to tokenizer loading regression." }, { "left": "issue:43618", "right": "issue:46285", "accept": false, - "reason": "Different CLIP-related regressions: attentions handling vs inconsistent model structure." + "reason": "Both are CLIP-related, but the concrete structural changes differ and do not point to one shared bug." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "Different implementation bugs: DSA indexer ReLU vs DeepSeekV3 MoE divergence." + "reason": "DSA indexer/ReLU discussion is unrelated to DeepSeekV3 implementation divergence." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "Test seed flakiness is unrelated to loading Phi-4 multimodal models." + "reason": "Test-seed flakiness and multimodal model loading failure are different issues." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "Different flaky-test causes: random-seed omission vs stale device override on CI." + "reason": "Both are CI/test-related, but one is about missing set_seed() and the other about a stale device override test." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "Local-files-only API requests and AddedToken keyword duplication are separate bugs." + "reason": "Offline-loading network leakage and AddedToken keyword duplication are unrelated bugs." }, { "left": "issue:42898", "right": "issue:46664", "accept": false, - "reason": "Tokenization cleanup behavior and video num_frames semantics are unrelated changes." + "reason": "Tokenization cleanup semantics and video num_frames semantics are different areas and different failure modes." }, { "left": "issue:44734", "right": "issue:46617", "accept": false, - "reason": "KV-cache continuation tensor indexing is a different code path from trust_remote_code cache FileNotFoundError." + "reason": "KV-cache continuation tensor-indexing crash and trust_remote_code cache FileNotFoundError are different code paths." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Tokenizer loading regression and checkpoint reload regression are similar only at a high level, not the same bug." + "reason": "Both are loading regressions, but tokenizer loading in v5 and checkpoint reload compatibility after version upgrade are not the same bug." }, { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "FP8 save_pretrained for MoE is a different issue from missing TP plan on Qwen3.5 MoE." + "reason": "FP8 save_pretrained behavior and missing tensor-parallel plan are distinct model concerns." }, { "left": "issue:43502", "right": "issue:44485", "accept": false, - "reason": "Offline loading requests and GLM-5 RoPE implementation are unrelated." + "reason": "Local-files-only network requests and GLM-5 RoPE implementation discussion are unrelated." }, { "left": "issue:43824", "right": "issue:46617", "accept": false, - "reason": "Missing export for a Qwen2.5-VL class is unrelated to loading a cached trust_remote_code model." + "reason": "Missing exported class import and cache FileNotFoundError are different failure modes." }, { "left": "issue:43874", "right": "issue:46617", "accept": false, - "reason": "Missing image-patch helper and cached remote-code FileNotFoundError are different failures." - }, - { - "left": "issue:43010", - "right": "issue:43065", - "accept": false, - "reason": "Cache update no_grad semantics do not match the Sam3PixelDecoder dummy Conv2d issue." + "reason": "Missing image-processor method and cache FileNotFoundError are unrelated." }, { "left": "issue:43576", "right": "issue:46132", "accept": false, - "reason": "Broken `transformers env` CLI and AttentionInterface.register behavior change are different problems." + "reason": "Broken transformers env command and AttentionInterface.register behavior changes are different issues." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Int4 quantization memory growth is not the same as failing to reload old checkpoints after upgrade." + "reason": "Int4 quantization memory growth/OOM and reload-after-upgrade failure are separate problems." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLflow reporting on failed training jobs is unrelated to tokenizer-loading regression." + "reason": "MLFlow reporting on failed training is unrelated to tokenizer loading regression." }, { "left": "issue:45125", "right": "issue:45478", - "accept": true, - "reason": "Same underlying Qwen3.5 MoE load path: one reports the missing `_tp_plan` root cause, the other reports the resulting `from_pretrained` failure." + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is missing _tp_plan and the other is a from_pretrained error; not the same underlying bug." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The only accepted pair, issue:45125 and issue:45478, is plausibly the same Qwen3.5 MoE load-path bug and looks mergeable into one fix. The other pair verdicts are reasonably cautious and do not overstate duplicates from the titles alone." + "accept": false, + "feedback": "Reject: the analyst overstates the only accepted soft pair. In the packet, issue:44393 (Qwen3-VL 2D bounding box hallucination/error) and issue:47000 (RTDetrLoss cost-matrix infeasible crash) are different model families and failure modes, and the packet itself marks the pair as deterministic_accept=false. The summary also says \u2018only one soft pair looks like the same underlying bug,\u2019 which is not grounded by the packet since none of the soft pairs are supported as mergeable." }, "evaluator_used": true, - "retried": false, - "accepted_nontrivial_soft_edge": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -18125,8 +18087,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -18215,20 +18177,20 @@ "issue:44485|issue:46664", "issue:44273|issue:45542" ], - "prepared_review_unit_hash": "5c7782725feab875ec235b849f1f9145c99ffe2ee9f74d61a0fd71ad90fe42b3", + "prepared_review_unit_hash": "033262e1e006fb94563d2a2e13fa292b045d92d494ce67409285d070e690586c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37263, + "serialized_chars": 37262, "estimated_input_tokens": 9316, "estimated_eval_tokens": 18888 }, @@ -18237,216 +18199,216 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster does not form a true duplicate set: the soft pairs are mostly loose topical similarities across unrelated issues (imports, loading regressions, docs, CLI, cache/generation). I rejected all soft edges and treated the cluster as heterogeneous rather than one underlying bug.", - "confidence": 0.89, - "canonical_issue_reason": "No single issue is a clear canonical duplicate target; the candidates span many distinct bugs and regressions, so a null canonical issue is more accurate than forcing a merge.", + "summary": "This cluster looks heterogeneous: most soft-similarity pairs are cross-topic false positives spanning unrelated import errors, model-loading regressions, docs issues, and model-specific bugs. I would not merge any of the proposed soft edges.", + "confidence": 0.86, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43576 is the closest representative only because it sits in the most soft-similarity links, but those links are still just broad CLI/dependency breakage similarity rather than duplicates.", + "best_issue_reason": "Issue 45850 is the broadest umbrella ('Since 5.0 version it breaks too many models'), but it is still too vague to serve as a strong canonical duplicate target for this mixed set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Different missing exports ('PreTrainedModel' vs 'HybridCache'); separate import breakages." + "reason": "Both are import errors, but for different symbols and likely different breakages (`PreTrainedModel` vs `HybridCache`)." }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled-Python execution errors vs checkpoint reload slowdown/regression; unrelated failures." + "reason": "One is about compiled Python files, the other about a training slowdown after upgrade; different bugs." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Loading hang vs documentation confusion; no shared underlying bug." + "reason": "Stuck loading is a runtime problem; the other is documentation confusion about `forward` arguments." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Tensor-parallel plan missing vs config field dropped by strict parsing; different Qwen issues." + "reason": "Tensor-parallel `_tp_plan` support and a missing config field are different model integration issues." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "local_files_only API requests vs FP8 save_pretrained bug; different codepaths." + "reason": "`local_files_only` network leakage and FP8 `save_pretrained` are unrelated behaviors." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "API request leakage vs unrelated placeholder issue; no meaningful overlap." + "reason": "No shared underlying bug; `racoon` is unrelated to offline loading." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs about removed pipelines vs unrelated placeholder issue; not the same bug." + "reason": "Docs still referencing `pipeline()` vs an unrelated `racoon` issue." }, { "left": "issue:46664", "right": "issue:46726", "accept": false, - "reason": "Video num_frames semantics vs missing-vision-backend handling; different processor bugs." + "reason": "Different video/image processor semantics; not the same bug." }, { "left": "issue:43010", "right": "issue:43232", "accept": false, - "reason": "Cache/layer update no_grad decoration vs generation sync_gpus behavior; different fixes." + "reason": "`no_grad` decoration on cache updates and generation kwargs handling are distinct generation concerns." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLflow job-status reporting vs tensor-parallel training bug; unrelated." + "reason": "MLflow job status reporting and tensor-parallel training are unrelated." }, { "left": "issue:29870", "right": "issue:31795", "accept": false, - "reason": "BART generate UnboundLocalError vs forward-docs confusion; not the same bug." + "reason": "Generation-time `decoder_attention_mask` crash is not the same as forward-doc confusion." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Both are version-related load regressions, but they describe different concrete loading failures." + "reason": "Both mention v5/v5.x regressions, but one is Qwen3-VL loading and the other is checkpoint reload compatibility; not the same concrete bug." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer loading regression vs a specific model-loading breakage; too different to merge." + "reason": "Both are tokenizer/model loading regressions, but for different model families and failure modes." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled-file runtime errors vs PretrainedConfig type-checking breakage; unrelated." + "reason": "Compiled-file execution errors and `PretrainedConfig` type-checking breakage are unrelated." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Missing huggingface_hub symbol vs broken transformers env CLI; different dependency and codepaths." + "reason": "Both are import/CLI-related v5 issues, but one is Hub import resolution and the other is the `transformers env` command." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "CLI breakage vs PretrainedConfig typing issue; not the same underlying problem." + "reason": "CLI breakage and config type-checking are different areas and not mergeable as one fix." }, { "left": "issue:43576", "right": "issue:47059", "accept": false, - "reason": "Both are CLI regressions, but one is transformers env and the other is a huggingface_hub TyperGroup incompatibility." + "reason": "`transformers env` CLI regression and unrelated huggingface_hub completion breakage are different code paths." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Model.forward docs issue vs missing configuration_utils export; unrelated." + "reason": "Documentation confusion in model.forward is not the same as a missing `layer_type_validation` import." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger-kernel model_init behavior vs a DSA indexer/ReLU discussion; different subsystems." + "reason": "Liger Kernel not applied in `model_init` and a DSA indexer discussion are unrelated." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "CLI breakage vs Qwen2.5-VL import error; only broad v5 breakage similarity." + "reason": "CLI breakage and missing `Qwen2_5_VLForConditionalGeneration` import are different regressions." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "Placeholder issue vs Gemma-4 audio positional encoding bug; no relation." + "reason": "`racoon` is unrelated to Gemma-4 audio positional encoding." }, { "left": "issue:44485", "right": "issue:44961", "accept": false, - "reason": "GLM-5 RoPE implementation vs placeholder issue; unrelated." + "reason": "GLM-5 RoPE implementation and `racoon` are unrelated." }, { "left": "issue:44351", "right": "issue:46617", "accept": false, - "reason": "ImportError for HybridCache vs FileNotFoundError when loading from cache; different causes and fixes." + "reason": "Both are cache/loading-related import problems, but they concern different missing behaviors and likely different modules." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor-parallel training bug vs docs about return_all_scores; unrelated." + "reason": "Tensor-parallel training and text-classification docs behavior are unrelated." }, { "left": "issue:43352", "right": "issue:44961", "accept": false, - "reason": "Flash Attention support error vs placeholder issue; no overlap." + "reason": "Flash Attention support for Nemotron and `racoon` share no underlying bug." }, { "left": "issue:43475", "right": "issue:46617", "accept": false, - "reason": "SAM 3 Video attribute error vs trust_remote_code cache FileNotFoundError; different load paths." + "reason": "SAM 3 Video attribute error and trust-remote-code cache loading are different failures." }, { "left": "issue:46143", "right": "issue:46144", "accept": false, - "reason": "Both concern RoFormer behavior, but one is kwargs propagation and the other is attention-interface usage; related area, not the same concrete bug." + "reason": "Missing `**kwargs` pass-through and attention-interface usage are related RoFormer topics but not the same concrete bug." }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "FP8 save_pretrained bug vs FP8 inference support request; different operations." + "reason": "FP8 save/load bug and FP8 inference support request are different problems." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "Different missing imports ('AutoImageProcessor' vs 'PreTrainedModel'); separate regressions." + "reason": "Missing `AutoImageProcessor` and missing `PreTrainedModel` imports are distinct export regressions." }, { "left": "issue:44315", "right": "issue:46209", "accept": false, - "reason": "Liger-kernel model_init bug vs FP8 inference on B200; unrelated." + "reason": "Liger Kernel application in `model_init` and FP8 inference on B200 are unrelated." }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Specific Qwen3.5 crash vs broad complaint that v5 breaks many models; not one concrete bug." + "reason": "Qwen3.5-35B chat crash is model-specific; the other is a broad v5 breakage complaint." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Generation cache missing vs cache-strategy typo; different issues." + "reason": "Generation cache missing during chunked prefill and a cache-strategy typo are not the same bug." }, { "left": "issue:44485", "right": "issue:46664", "accept": false, - "reason": "GLM-5 RoPE implementation vs SmolVLM num_frames semantics; unrelated model areas." + "reason": "GLM-5 RoPE implementation and SmolVLM frame-count semantics are unrelated." }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Lazy loading problem vs tensorboard/TensorFlow backend error; no shared codepath." + "reason": "Lazy loading failure and undefined TensorFlow backend error are different issues." } ] }, @@ -18465,8 +18427,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -18560,275 +18522,272 @@ "issue:44273|issue:44908", "issue:45230|issue:45310" ], - "prepared_review_unit_hash": "14437dd61165a63c8f35136db60e2823e988720a4da070b5a733771e8352426d", + "prepared_review_unit_hash": "0eed20966a3a48c028dfb7d9e2d220e528936e20300fd57c387c3d7951c577a3", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 39, - "serialized_chars": 38826, - "estimated_input_tokens": 9707, - "estimated_eval_tokens": 19670 + "serialized_chars": 38824, + "estimated_input_tokens": 9706, + "estimated_eval_tokens": 19668 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of Transformers v5 regression reports, import/load issues, and a few unrelated model-specific bugs. The overlap is usually only at the subsystem level, not the same concrete defect, so I would not merge any soft pairs.", - "confidence": 0.56, - "canonical_issue_reason": "issue:45850 is the broadest umbrella in the set: it explicitly captures the general \"v5 breaks too many models\" regression theme that many of the other issues orbit around.", + "summary": "This set is mostly a grab-bag of unrelated Transformers regressions and model-specific bugs. Only one soft duplicate looks plausible: the lazy-loading/import path issues around issue 44273 and issue 45003.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44273 is the broadest and most central report in the only coherent thread here (lazy loading/import behavior), and it has multiple nearby soft-similarity candidates.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45850 is the best global representative because it is the most general duplicate magnet for the recurring v5 breakage complaints in this cluster.", + "best_issue_reason": "Issue 44273 is the best issue anchor: open, discussion-heavy, and directly names the shared failure mode the closest duplicate candidate appears to target.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "45230 is too generic; 44623 is a specific processor.save_pretrained missing-files bug." + "reason": "'processor.save_pretrained(...) missing files' and a generic 'Bug report' are too underspecified to be the same bug." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "Both are Qwen3/Qwen3-VL related, but one is about model support in v5 and the other is about missing tensor-parallel metadata." + "reason": "Both involve Qwen3.5/Moe, but one is a general bug report and the other is a missing _tp_plan tensor-parallelism issue; different failure modes." }, { "left": "issue:44273", "right": "issue:45003", - "accept": false, - "reason": "Lazy-loading/import performance and unsafe sys.modules access are related areas, but they are not the same concrete bug." + "accept": true, + "reason": "Both point to lazy loading/import machinery failing; unsafe sys.modules access in modeling_utils is a plausible root cause for 'lazy loading is not working properly'." }, { "left": "issue:44360", "right": "issue:46664", "accept": false, - "reason": "These are unrelated bugs in completely different components." + "reason": "DSA indexer/ReLU discussion is unrelated to SmolVLM video processor frame-count semantics." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Import-star cleanup/performance and missing GenerationMixin cache in chunked_prefill are different problems." + "reason": "Import-* cleanup/performance and GenerationMixin cache chunked-prefill are different code paths and symptoms." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "A Qwen3-VL bounding-box/output bug is not the same as a PretrainedConfig type-checking regression." + "reason": "Qwen3-VL bbox output problems are unrelated to PretrainedConfig type checking regressions." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Same model family, but one is 2D box output failure and the other is an infinite-loop/image-specific generation bug." + "reason": "Same model family, but 2D bbox hallucinations and infinite loop on a specific image are distinct concrete bugs." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "CLI env-command breakage and checkpoint reload incompatibility are different failures." + "reason": "Broken 'transformers env' CLI and checkpoint reload failures after v4.57->v5.x are unrelated." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "Missing pad_token_id on Llama4Config is a specific config attribute bug, not the broad generic v5 breakage report." + "reason": "Missing pad_token_id on Llama4Config is a specific config bug, not the broad 'v5 breaks too many models' report." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both concern old checkpoints in v5, but they appear to be different model/code-path incompatibilities." + "reason": "Both are v5 upgrade regressions, but one is general checkpoint reload and the other is InternVL2 remote-code/meta-init incompatibility." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "StableLmConfig missing pad_token_idx is a specific attribute regression, not the umbrella issue." + "reason": "StableLm pad_token_idx regression is model-specific and not the same as the broad version-5 breakage complaint." }, { "left": "issue:44492", "right": "issue:45850", "accept": false, - "reason": "A typo in cache-strategy docs/labels is unrelated to broad model breakage." + "reason": "A cache-strategy typo is unrelated to generic v5 model breakage." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 inference support and a weight-conversion question are not the same underlying bug/change." + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5 Moe is not the same as a weight conversion-script question." }, { "left": "issue:45850", "right": "issue:47059", "accept": false, - "reason": "Generic v5 breakage and a huggingface_hub CLI test failure are different issues." + "reason": "Broad v5 model breakage and a cli test failure with huggingface_hub 1.22.0 are unrelated." }, { "left": "issue:45278", "right": "issue:45850", "accept": false, - "reason": "Both are broad v5-regression complaints, but 45278 is specifically about import errors and not the same underlying defect as 45850." + "reason": "Both mention many breakages after upgrade, but they are different umbrella reports rather than the same underlying bug." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API guidance and a cache-strategy typo are unrelated." + "reason": "TF32 API guidance and a cache-strategies typo are unrelated." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import-time slowdown from import-star usage is not the same as a missing Qwen3_5 module import." + "reason": "Import-time slowdown from import-* usage and a specific Qwen3_5 module import failure are different problems." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model-type warning and Liger Kernel application behavior are unrelated." + "reason": "sam3_video/sam3_tracker warning and Liger Kernel not applied with model_init are unrelated." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Both mention import/lazy loading, but one is intermittent import slowness and the other is lazy loading not working properly; too broad to merge." + "reason": "'import transformers takes long sometimes' is a performance symptom, but not clearly the same concrete bug as lazy loading not working properly." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "A flaky test due to missing set_seed calls is unrelated to inverse_sqrt scheduler kwargs handling." + "reason": "Missing set_seed in tests and inverse_sqrt scheduler kwargs handling are unrelated." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "CLI env breakage and scheduler kwargs are different subsystems and different bugs." + "reason": "Broken env CLI and scheduler kwargs behavior are different subsystems." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash-Attention support gating and Qwen3-VL bounding-box hallucinations are unrelated." + "reason": "Flash Attention 2 unsupported on NemotronH and Qwen3-VL bbox hallucinations are unrelated." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version compatibility and a Qwen3-VL output bug are unrelated." + "reason": "Python version compatibility and Qwen3-VL output errors are unrelated." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "A scheduler kwargs bug and a PretrainedConfig typing regression are different code paths." + "reason": "Scheduler kwargs bug and PretrainedConfig type checking regression are unrelated." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "Registering position_ids as a buffer causing read errors is not the same as the generic bug report." + "reason": "position_ids buffer access error and a generic bug report are not the same issue." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer save_pretrained class mismatch is a specific serialization bug, not the generic bug report." + "reason": "tokenizer.save_pretrained class-name mismatch is unrelated to a generic bug report." }, { "left": "issue:44492", "right": "issue:46328", "accept": false, - "reason": "Cache-strategy typo and SequenceFeatureExtractor.pad performance are unrelated." + "reason": "Cache strategy typo and SequenceFeatureExtractor.pad inefficiency are unrelated." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-star performance work and missing HybridCache export are different import-related problems." + "reason": "import-* cleanup/perf and missing HybridCache import are different import-path bugs." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor cached_file kwargs forwarding and scheduler kwargs handling are unrelated." + "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs are different APIs and code paths." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import-star cleanup/performance and a testing_utils bug are unrelated." + "reason": "Model import performance and a testing_utils bug are unrelated." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "Python-version compatibility and a missing Qwen3_5 module import are different problems." + "reason": "Python compatibility and Qwen3_5 import failure are unrelated." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init policy for patchtsmixer and _get_tied_weight_keys crashing on a list are unrelated." + "reason": "patchtsmixer post_init policy and _get_tied_weight_keys crashing on list-typed keys are not the same bug." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion and tokenizer-loading regression are not the same underlying bug." + "reason": "Documentation confusion in model.forward and tokenizer loading regression are unrelated." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "Broken env command and lazy loading not working are both tooling/import-adjacent, but not the same concrete defect." + "reason": "Broken CLI env command and lazy loading not working are different failures." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "Liger Kernel not being applied in model_init and flash_attention_3 causing degenerate training are different issues." + "reason": "Liger Kernel application in model_init and flash_attention_3 degenerate training are different bugs." }, { "left": "issue:43519", "right": "issue:46655", "accept": false, - "reason": "Both are video-processor bugs, but timestamp calculation and padding calculation are separate defects." + "reason": "Incorrect timestamp calculation and incorrect video padding calculation are different processor bugs." }, { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "DSA indexer/ReLU discussion and Gemma-4 audio positional encoding are unrelated." - }, - { - "left": "issue:44273", - "right": "issue:44908", - "accept": false, - "reason": "Lazy loading/import behavior and scheduler kwargs handling are unrelated." + "reason": "DSA indexer/ReLU discussion is unrelated to Gemma4AudioRelPositionalEncoding." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "Generic bug report and Qwen3.5 Moe from_pretrained error are not the same issue." + "reason": "Generic bug report and Qwen3.5 Moe from_pretrained error are not the same concrete issue." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft edge (44273 \u2194 45003) is framed as plausible rather than certain, and the other rejections are appropriately cautious and specific to distinct failure modes." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -18840,8 +18799,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -18938,286 +18897,286 @@ "issue:42936|issue:46018", "issue:43668|issue:44031" ], - "prepared_review_unit_hash": "5c330680caa6f90524af34e66f8285f4820e4cfd263e792b3055d5162aed01a0", + "prepared_review_unit_hash": "006cc7982c7a59fe9df36a81f5d08547ea14176b607ed55fba5169ee3b88ff4c", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 42, - "serialized_chars": 39471, - "estimated_input_tokens": 9868, - "estimated_eval_tokens": 19992 + "serialized_chars": 39464, + "estimated_input_tokens": 9866, + "estimated_eval_tokens": 19988 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous. The soft pairs share superficial vocabulary (loading, tokenizer, Flash Attention, positional encodings, etc.) but do not look like the same underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue fits well: the items span unrelated regressions, feature requests, docs, and import/runtime failures.", + "summary": "This set is highly heterogeneous: most soft pairs are only title-similar but concern different models, subsystems, or release regressions. No PRs are present. If a single umbrella issue must be chosen, the broad v5-breakage report is the closest fit, but it is not a true duplicate target for most items.", + "confidence": 0.44, + "canonical_issue_reason": "issue:45850 is the broadest umbrella in the set, describing many model breakages after v5; if one issue must represent the cluster, it is the closest fit.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue; none is representative of the cluster as a whole beyond broad subsystem overlap.", + "best_issue_reason": "issue:45850 is the most representative high-level regression report; the other issues are mostly narrower, unrelated bugs or feature requests.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Both involve sequence-structure math, but they target different models and different bugs." + "reason": "Different code paths: Qwen3VL timestamping vs GLM-5 RoPE implementation." }, { "left": "issue:45468", "right": "issue:46894", "accept": false, - "reason": "Audio-related, but one is a positional-encoding issue and the other is a training-loss shift bug." + "reason": "Different models and bugs: Gemma audio positional encoding vs CohereASR training-loss shifting." }, { "left": "issue:43576", "right": "issue:46218", "accept": false, - "reason": "CLI/env command failure and model import failure are unrelated code paths." + "reason": "Broken env CLI vs Qwen3_5 import failure are unrelated." }, { "left": "issue:44655", "right": "issue:44861", "accept": false, - "reason": "Both touch saving, but one is pipeline serialization and the other is tied-weight key handling." + "reason": "Saving Pipeline objects and tied-weight-key handling are separate problems." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Different components: processor timestamp calculation vs scheduler kwargs handling." + "reason": "Timestamp calculation bug is unrelated to inverse_sqrt scheduler kwargs handling." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import cleanup/performance and multi-thread VRAM leakage are unrelated problems." + "reason": "Import-star cleanup performance issue is unrelated to a VRAM leak." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Docs confusion versus a concrete model-loading regression." + "reason": "Documentation confusion about forward args is not the same as a Llama model loading regression." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "Both are positional-encoding themed, but for different models and different failures." + "reason": "Different model families and different implementation areas." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "Concurrency VRAM leak is unrelated to RoPE implementation details." + "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to GLM-5 RoPE behavior." }, { "left": "issue:46018", "right": "issue:46172", "accept": false, - "reason": "Same model family, but different concrete bugs: expert gating versus tensor transpose/casting." + "reason": "Shared-expert gating bug and DeepSeek weight-transpose issue are different code paths." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Local-files-only network leakage is unrelated to timestamp miscalculation." + "reason": "Unexpected network requests from local_files_only is unrelated to Qwen3VL timestamp math." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 configuration and VRAM leakage are different subsystems and symptoms." + "reason": "TF32 API guidance and a dataloader-thread VRAM leak are unrelated." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "A broad complaint about 5.0 breakage is not the same as a specific import error." + "reason": "Broad v5 regression claim and a specific Qwen3_5 import failure are not the same bug." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Processor timestamp math and save_pretrained missing files are different bugs." + "reason": "Processor timestamping and processor.save_pretrained missing files are different issues." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "Different model code paths and different failure modes." + "reason": "torch.split return values in GLM-MoE indexing is unrelated to Gemma audio positional encoding." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "Both mention normalization eps, but they affect different models and different logic." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is not the same as Gemma audio positional encoding." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Continuous batching support and weight-conversion tooling are unrelated." + "reason": "Audio continuous batching support and DeepSeek-V4 weight-conversion questions are unrelated." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Timestamp calculation and VRAM leakage do not share a concrete code path." + "reason": "Qwen3VL timestamp calculation and Qwen3ForCausalLM VRAM leakage are separate bugs." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "Both are import errors, but for different symbols/modules and different causes." + "reason": "Different import failures for different symbols/models; not the same concrete failure." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Flash-Attention support and an import failure are distinct issues." + "reason": "Nemotron Flash Attention support and Qwen3_5 import failure are unrelated." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Both mention tokenizer v5 behavior, but the reported symptoms and likely fixes differ." + "reason": "Both mention tokenizer v5, but one is behavior drift and the other is decoder-type mismatch; not the same code-path bug." }, { "left": "issue:44861", "right": "issue:47059", "accept": false, - "reason": "Serialization helper crash and CLI completion regression are unrelated." + "reason": "Tied-weight-key crash and CLI completion breakage are unrelated." }, { "left": "issue:28348", "right": "issue:45250", "accept": false, - "reason": "One is GPT2-specific support; the other is a vague generic Flash Attention request." + "reason": "GPT2-specific Flash Attention 2.0 support and a generic FA2 request are not the same concrete change." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Scheduler kwargs bug and spam/off-topic issue are unrelated." + "reason": "Scheduler kwargs bug is unrelated to the placeholder issue titled 'Unemployment'." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "A cache-strategy typo and an unrelated spam issue are not duplicates." + "reason": "Cache-strategy typo and an unrelated placeholder issue are not duplicates." }, { "left": "issue:44998", "right": "issue:45850", "accept": false, - "reason": "Generic breakage complaint versus unrelated off-topic issue." + "reason": "The placeholder issue is unrelated to v5-wide model breakage." }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attentions regression is unrelated to the off-topic issue." + "reason": "CLIPOutput attentions regression is unrelated to the placeholder issue." }, { "left": "issue:44938", "right": "issue:44998", "accept": false, - "reason": "Python 3.14 compatibility issue is unrelated to the off-topic issue." + "reason": "Python 3.14 loading failure is unrelated to the placeholder issue." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weights crash and chat-template tool-call crash are different code paths." + "reason": "Tied-weight-key crash and assistant tool-call template crash are different bugs." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Checkpoint loading performance and import error are unrelated." + "reason": "Slow from_pretrained loading and an ImportError from configuration_utils are unrelated." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "Config field missing and tied-weight helper crash are different bugs." + "reason": "StableLmConfig pad_token_idx regression is unrelated to tied-weight-key handling." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 support request and processor save_pretrained missing files are unrelated." + "reason": "FP8 Qwen3-VL/Qwen3.5 support request and processor.save_pretrained missing files are different issues." }, { "left": "issue:28348", "right": "issue:43352", "accept": false, - "reason": "Both are Flash Attention 2.0 support requests, but for different models and not one concrete fix." + "reason": "Both are Flash Attention 2.0 related, but for different models and different concrete support gaps." }, { "left": "issue:29870", "right": "issue:31515", "accept": false, - "reason": "Generation-time decoder mask crash and slow loading are unrelated." + "reason": "BART generate UnboundLocalError and slow checkpoint loading are unrelated." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow reporting failure and checkpoint loading speed are different issues." + "reason": "MLflow reporting on failure and slow from_pretrained loading are unrelated." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Loading performance and tensor-parallel training bug do not share a concrete fix." + "reason": "Checkpoint load slowness and tensor-parallel training bugs do not share the same code path." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Performance bug and documentation confusion are not duplicates." + "reason": "Loading performance and forward-argument documentation confusion are unrelated." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "FP8 support request and model from_pretrained regression are different failures." + "reason": "FP8 inference support and Qwen3.5 from_pretrained loading error are different issues." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Continuous batching support and FP8 inference support are separate feature requests." + "reason": "Audio continuous batching and Qwen3-VL/Qwen3.5 FP8 support are unrelated." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Different architectures and different implementation problems." + "reason": "Mamba init_weights behavior and RoFormer attention-interface usage are different problems." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2former ignore_value misuse and DeepSeek expert gating are unrelated." + "reason": "Mask2Former ignore_value handling is unrelated to DeepSeek-V4 shared-expert gating." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "A config type hint issue and a tokenizer regex warning are unrelated." + "reason": "Config type-hint correction and tokenizer regex warning regression are unrelated." } ] }, @@ -19236,8 +19195,8 @@ "review_unit_count": 62, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-509", "nodes": [ @@ -19260,14 +19219,14 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "2f78c3d88601856e8412d5114d8f2816cd6cbab49d397b9d5660729d208e0b03", + "prepared_review_unit_hash": "4c2eb0795ea6e7ffc424e6eb4fa9751ea806f696ded517e3b1b7af2447933ca6", "component_budget": { "node_count": 509, "item_count": 509, - "soft_pair_count": 1869, - "serialized_chars": 970140, - "estimated_input_tokens": 242535, - "estimated_eval_tokens": 485326 + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 }, "budget": { "node_count": 7, @@ -19282,66 +19241,66 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are about different models, configs, tokenizer warnings, and a typo, so they do not look like duplicates of one underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": "No true canonical issue stands out: the cluster is heterogeneous and spans unrelated code paths, so any single issue would be an arbitrary representative rather than a duplicate hub.", + "summary": "This cluster is heterogeneous: the issues cover unrelated models, a config type hint, a typo, tokenizer warnings, and a separate attention-interface bug. None of the soft pairs look like the same underlying defect, so they should not be merged as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "If forced to anchor the cluster, issue 43240 is the clearest concrete bug report, but it is not actually duplicated by the others; most other items target different components or different classes of problems.", "canonical_pr_reason": null, - "best_issue_reason": "If forced to pick one representative, #43240 is the broadest runtime bug report, but it still does not explain the rest of the cluster and should not be treated as a canonical duplicate.", + "best_issue_reason": "Issue 43240 is the most actionable and specific runtime bug report in the set, with clear symptom language and extra discussion/inbound references, making it the best standalone representative even though the cluster is not a duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Different bugs: Mask2Former ignore_value handling vs kwargs not being forwarded into loss calculation." + "reason": "Different bugs in different models: Mask2Former ignore_value handling vs. kwargs not forwarded to a loss function." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Unrelated code paths: loss-function kwargs plumbing vs RoFormer attention interface usage." + "reason": "Unrelated code paths: loss-function kwargs forwarding versus RoFormer attention not using the attention interface." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Segmentation loss parameter usage and tokenizer regex warnings are unrelated issues." + "reason": "Mask2Former loss-parameter usage and tokenizer regex warning are unrelated subsystems and symptoms." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Mask2Former ignore_value handling does not overlap with RoFormer attention interface behavior." + "reason": "Different model bugs with no shared concrete code-path; one is loss masking, the other is attention implementation structure." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "One is a model-loss parameter bug; the other is a typo in a quantization attribute name." + "reason": "A model behavior bug versus a typo in a quantization-related attribute name; not the same underlying issue." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "ModernBERTConfig type-hint correctness is unrelated to a string typo in quantization conversion code." + "reason": "Config type-hint correctness and a spelling/attribute typo are distinct, non-overlapping defects." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwargs forwarding and config type-hint correctness are different classes of problems." + "reason": "Loss-function kwargs forwarding is a runtime behavior bug; ModernBERTConfig norm_eps typing is a separate metadata/type issue." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Llama norm implementation question and ModernBERT config typing issue do not share an underlying fix." + "reason": "Llama4 normalization implementation question and ModernBERT config type hint incorrectness are unrelated topics." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value behavior is unrelated to ModernBERT norm_eps typing." + "reason": "Different models and problem types: Mask2Former ignore_value usage versus ModernBERT config typing." } ] }, @@ -19360,8 +19319,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10519-3", "nodes": [ @@ -19374,7 +19333,7 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "ffb898021582bce1148e81abf366879935dc8bd93df67948963ef365ae46aa1e", + "prepared_review_unit_hash": "620985b08c09c133bf067c63933fe9b833ae37f17210786825eab26c1a46e41c", "component_budget": { "node_count": 3, "item_count": 3, @@ -19396,30 +19355,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The three issues are not duplicates: they cover token truncation behavior, ReduceLROnPlateau-style scheduler functionality, and Wav2Vec2 word offset timing. No soft edge should be accepted.", + "summary": "These three issues are not duplicates: one requests tokenizer truncation-from-start behavior, one asks for ReduceLROnPlateau-like training logic, and one concerns Wav2Vec2 word boundary timestamps. They are separate feature requests in different areas.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "No single canonical issue fits this cluster; the items describe unrelated feature requests rather than the same bug or change.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a true canonical for the cluster because the topics are unrelated. If forced to pick the most self-contained thread, issue #10519 is the clearest standalone request, but it is not representative of the others.", + "best_issue_reason": "Issue 10519 is the most detailed and clearly scoped of the three, but it is still not a canonical duplicate target because the topics are unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different domains: tokenizer truncation direction vs Wav2Vec2 word boundary timestamps. No shared underlying bug or change." + "reason": "Different requests: tokenizer truncation direction vs Wav2Vec2 word time offsets." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "One asks for truncation-from-beginning options; the other asks for ReduceLROnPlateau-like learning-rate scheduling. Unrelated feature requests." + "reason": "Different functionality: truncation behavior vs ReduceLROnPlateau-style scheduling." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "Scheduler functionality request vs audio model word-offset timing. Distinct subsystems and problems." + "reason": "Unrelated topics: learning-rate scheduling vs speech token boundary timestamps." } ] }, @@ -19438,8 +19397,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-13244-3", "nodes": [ @@ -19451,46 +19410,46 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "9f51535e1cc0ee73e9984ba39437c87fd46e21871d13ba8df11c31cb34e2926f", + "prepared_review_unit_hash": "16f5ebe336d70c9e95ee6e9061ef2e831443265b3b69ede2aa533065167dfd19", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2476, - "estimated_input_tokens": 619, - "estimated_eval_tokens": 1494 + "serialized_chars": 2479, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2476, - "estimated_input_tokens": 619, - "estimated_eval_tokens": 1494 + "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 three issues are not duplicates: one is about Tapas tokenization parity, one about an unhelpful auth/download error for private models, and one about failing to load bert-base-uncased. The only commonality is broad model/tokenizer loading context, not the same underlying bug.", - "confidence": 0.91, - "canonical_issue_reason": "issue #16618 is the broadest and most representative of the cluster because it concerns a concrete model-loading failure; the other two are separate problems in tokenization and download/auth messaging.", + "summary": "These issues are mostly not duplicates. Two are about model loading, but they point to different failure modes; the Tapas tokenization report is unrelated.", + "confidence": 0.82, + "canonical_issue_reason": "issue:16618 is the broadest and most general model-loading report in the set, so it is the best representative if a single issue must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "#16618 is the best issue to keep as the representative because it describes a direct user-facing failure in model loading and is closer to the shared high-level theme than the other two, which are unrelated specifics.", + "best_issue_reason": "issue:16618 has the clearest generic symptom wording and is the most plausible cluster anchor among the three.", "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. Same subsystem, but not the same bug or fix." + "reason": "Tapas tokenization vs. failing to load bert-base-uncased are different subsystems and different bug classes; no same underlying change or fix is apparent." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "The first is about misleading error text for private-model download authentication; the second is a generic model loading failure. These are distinct issues." + "reason": "Private-model download auth/error-message behavior is not the same as a public model load failure for bert-base-uncased; symptoms and likely code paths differ." } ] }, @@ -19509,8 +19468,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-26350-2", "nodes": [ @@ -19520,40 +19479,40 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "427b0a76097f8ee011b4726bab179e72b27e824b675e66cbb6728d012b5bad56", + "prepared_review_unit_hash": "78dd7e0ef3c51041f23c21ce5d6299bc07c572339903743979b0cd594d1670c8", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, - "estimated_input_tokens": 398, - "estimated_eval_tokens": 1052 + "serialized_chars": 1588, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, - "estimated_input_tokens": 398, - "estimated_eval_tokens": 1052 + "serialized_chars": 1588, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two distinct community feature requests: one for Flash Attention 2 support across more architectures, and one for GGUF support across more architectures. They share only broad wording and are not the same underlying change.", - "confidence": 0.98, - "canonical_issue_reason": "No single canonical issue fits: the requests target different features and code paths (Flash Attention 2 vs GGUF support).", + "summary": "These are two unrelated community-contribution feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They share only a broad theme of adding architecture support, not the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a true duplicate of the other; if one must be kept as representative, choose based on the larger current discussion, but not as a canonical match for this pair.", + "best_issue_reason": "Neither issue is a duplicate of the other; if one must be kept as a representative issue, there is no clear canonical choice because they request different features and code paths.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Reject: these are different feature requests for different model/runtime support paths, not the same bug or change." + "reason": "Different enhancement requests for different subsystems/formats (Flash Attention 2 vs GGUF); they do not describe the same bug or change and would not plausibly be merged into one issue." } ] }, @@ -19572,8 +19531,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-27088-9", "nodes": [ @@ -19618,215 +19577,218 @@ "issue:46222|issue:46931", "issue:46226|issue:46931" ], - "prepared_review_unit_hash": "521e3c32dc86761b90cf561fa0e1bfcbf8b0a8e4679880d27c2075a26e7b6d80", + "prepared_review_unit_hash": "7d8c929d9a4eaa5c25318a5855fd2f20d73dd423eb609764c8045d86633165b7", "component_budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 29, - "serialized_chars": 15975, - "estimated_input_tokens": 3994, - "estimated_eval_tokens": 8244 + "serialized_chars": 15980, + "estimated_input_tokens": 3995, + "estimated_eval_tokens": 8246 }, "budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 29, - "serialized_chars": 15975, - "estimated_input_tokens": 3994, - "estimated_eval_tokens": 8244 + "serialized_chars": 15980, + "estimated_input_tokens": 3995, + "estimated_eval_tokens": 8246 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are superficially similar i18n translation issues, but they describe separate language-specific translation tasks rather than the same underlying bug/change. I would not dedupe them; at most, issue 43834 is the cleanest representative template-style issue in the set.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 43834 is the most generic template-style translation issue and best represents the shared pattern, but the cluster does not contain a true duplicate set.", + "summary": "This cluster is mixed. The visible metadata supports one exact duplicate pair, 46222/46226, with identical Thai titles and matching body metadata. The other issues split into distinct explicit i18n titles ([i18n-TR], [i18n-es], [i18n-ro]) and several placeholder-title issues with no body/link metadata strong enough to merge on title similarity alone.", + "confidence": 0.87, + "canonical_issue_reason": "46222 is the best canonical issue for the only clearly duplicated subgroup because it is an exact match with 46226 and has inbound references.", "canonical_pr_reason": null, - "best_issue_reason": "43834 is the least language-specific and most representative of the repeated i18n-docs template issues.", + "best_issue_reason": "46222 is the strongest overall issue anchor for triage because it represents the only exact duplicate pair in the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different translation tasks; Turkish vs a generic/template language entry, so not the same concrete change." + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different translation tasks; Turkish vs a generic/template language entry, so not the same concrete change." + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Same broad docs-i18n theme, but one is a specific Turkish request and the other is a generic template issue." + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Turkish and Romanian translation requests are separate language-specific tasks, not duplicates." + "reason": "Different visible language-coded titles ([i18n-TR] vs [i18n-ro]); the metadata shows separate translation requests." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish and Spanish translation requests are separate language-specific tasks, not duplicates." + "reason": "Different visible language-coded titles ([i18n-TR] vs [i18n-es]); the metadata shows separate translation requests." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish-specific request versus a generic template issue; related category, but not the same task." + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Both are translation-template style issues, but there is no evidence they refer to the same concrete language task." + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:28936", - "right": "issue:44370", + "right": "issue:38435", "accept": false, - "reason": "Both are translation-template style issues, but there is no evidence they refer to the same concrete language task." + "reason": "Different visible language-coded titles ([i18n-es] vs [i18n-ro]); the metadata shows separate translation requests." }, { "left": "issue:28936", - "right": "issue:38435", + "right": "issue:44370", "accept": false, - "reason": "Spanish and Romanian are distinct translation requests, not duplicates." + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian-specific request versus a generic template issue; related category, but not the same task." + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Romanian-specific request versus a generic template issue; related category, but not the same task." + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Romanian-specific request versus a generic template issue; related category, but not the same task." + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." }, { "left": "issue:38435", "right": "issue:46222", "accept": false, - "reason": "Different language/request context; no evidence they are the same translation change." + "reason": "The visible titles are unrelated: a Romanian i18n request versus a Thai-titled issue; the packet shows no shared concrete request." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "Different language/request context; no evidence they are the same translation change." + "reason": "The visible titles are unrelated: a Romanian i18n request versus a Thai-titled issue; the packet shows no shared concrete request." }, { "left": "issue:43834", "right": "issue:44370", "accept": false, - "reason": "These look like repeated instances of the same translation template, but not enough evidence that they are one underlying issue." + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." }, { "left": "issue:43834", "right": "issue:44365", "accept": false, - "reason": "These look like repeated instances of the same translation template, but not enough evidence that they are one underlying issue." + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." }, { "left": "issue:43834", "right": "issue:46222", "accept": false, - "reason": "Shared i18n-docs pattern, but the titles/body signals are not enough to prove a single duplicate task." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Shared i18n-docs pattern, but the titles/body signals are not enough to prove a single duplicate task." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:43834", "right": "issue:46931", "accept": false, - "reason": "The Thai-garbled issue appears to be a separate request, not the same concrete translation task as the generic template issue." + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." }, { "left": "issue:44365", "right": "issue:44370", "accept": false, - "reason": "Very similar template issues, but no concrete evidence they are the same underlying request." + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Similar i18n template wording, but not enough to conclude a single duplicate issue." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Similar i18n template wording, but not enough to conclude a single duplicate issue." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:44365", "right": "issue:46931", "accept": false, - "reason": "Different-looking translation issue; no clear evidence of the same underlying task." + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Similar i18n template wording, but not enough to conclude a single duplicate issue." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Similar i18n template wording, but not enough to conclude a single duplicate issue." + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." }, { "left": "issue:44370", "right": "issue:46931", "accept": false, - "reason": "Different-looking translation issue; no clear evidence of the same underlying task." + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." }, { "left": "issue:46222", "right": "issue:46226", - "accept": false, - "reason": "Identical-looking titles, but still not enough concrete evidence that these are the same underlying request rather than separate instances." + "accept": true, + "reason": "Exact duplicate: the Thai titles and visible body metadata match, and 46222 has inbound references." }, { "left": "issue:46222", "right": "issue:46931", "accept": false, - "reason": "The titles suggest separate language/text variants, not a single duplicate bug or change." + "reason": "Different Thai titles with different visible text, so they are not the same issue." }, { "left": "issue:46226", "right": "issue:46931", "accept": false, - "reason": "The titles suggest separate language/text variants, not a single duplicate bug or change." + "reason": "Different Thai titles with different visible text, so they are not the same issue." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The exact duplicate call on 46222/46226 is grounded, but the broader summary and several soft-edge reasons go beyond the packet. In particular, 43834/44365/44370 all have the same placeholder title and only soft-similarity evidence; the packet does not show they are different language requests, so that part is overconfident. Please keep the cluster description and pair reasons strictly tied to the visible titles/body metadata." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -19838,8 +19800,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-36718-21", "nodes": [ @@ -19915,322 +19877,304 @@ "issue:42503|issue:46653", "issue:45027|issue:46653" ], - "prepared_review_unit_hash": "92e65a7163589ad659cac0d9a40063efcd6d0b9427a6c537656e00d95049ca6e", + "prepared_review_unit_hash": "260743dceed40fee5e79a496bd50cd62850da40cb93f0ee662395e8db5c37bfe", "component_budget": { "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28613, - "estimated_input_tokens": 7154, - "estimated_eval_tokens": 14564 + "serialized_chars": 28621, + "estimated_input_tokens": 7156, + "estimated_eval_tokens": 14568 }, "budget": { "node_count": 21, "item_count": 21, "soft_pair_count": 48, - "serialized_chars": 28613, - "estimated_input_tokens": 7154, - "estimated_eval_tokens": 14564 + "serialized_chars": 28621, + "estimated_input_tokens": 7156, + "estimated_eval_tokens": 14568 }, "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-add/config requests across vision, text, audio, and OCR. The shared wording (\u201cAdd/Support/Proposal\u201d) drives soft similarity, but the underlying changes are different, so no pair looks like a true duplicate.", - "confidence": 0.97, - "canonical_issue_reason": "No clear canonical issue: the items span distinct model requests and config changes rather than one duplicated problem.", + "summary": "This cluster is only loosely related by topic (mostly model-addition/support requests); the items are not duplicate-level matches and span different models, modalities, and use cases. No PRs are present.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43837 is the most central concrete support request in the similarity graph, but this is not a true duplicate cluster; the other items are mostly distinct model requests.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a safe global representative; the cluster is heterogeneous across model families and tasks.", + "best_issue_reason": "issue:43837 is the best single anchor only because it is a clear, specific model-support request with many soft links; however, the linked issues are still different changes, so it should not be treated as a duplicate canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36718", "right": "issue:47004", "accept": false, - "reason": "Same general keypoint domain, but RoMa matcher vs LoMa detector+matcher are different concrete requests." + "reason": "Both are keypoint matcher requests, but RoMa vs LoMa are different models and changes; not the same underlying feature." }, { "left": "issue:36718", "right": "issue:41115", "accept": false, - "reason": "Keypoint matcher request vs MiniCPM3 model architecture request; unrelated." + "reason": "Keypoint matching vs MiniCPM3 architecture are unrelated model requests." }, { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config addition vs DEIMv2 model addition; different model families and changes." + "reason": "EuroBert config addition and DEIMv2 support are different model entries." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "Text-model config request vs Qwen3-TTS speech support; unrelated." + "reason": "EuroBert config work and Qwen3-TTS support are different model families and modalities." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert config work and RF-DETR model work are different concrete requests." + "reason": "EuroBert and RF-DETR are unrelated models; same broad area only." }, { "left": "issue:36817", "right": "issue:42954", "accept": false, - "reason": "EuroBert config vs MiMo model request; no shared code-path issue." + "reason": "EuroBert config and XiaoMi MiMo are different model support requests." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "Both are detection models, but RF-DETR and DEIMv2 are distinct models with separate implementations." + "reason": "RF-DETR and DEIMv2 are different detection models, not one concrete change." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "Object-detection model vs audio model request; unrelated." + "reason": "RF-DETR and LFM2.5 Audio 1.5B are unrelated model additions." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR model addition vs weight-sparse transformer request; different architecture and scope." + "reason": "RF-DETR vs a weight-sparse transformer request are different code paths and model types." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "Detection model request vs Qwen3-ASR speech support; different tasks." + "reason": "RF-DETR and Qwen3-ASR are different models and different support work." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "Detection model request vs Qwen3-TTS speech support; different tasks." + "reason": "RF-DETR and Qwen3-TTS are unrelated features." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT text-model request vs Qwen3-ASR speech support; unrelated." + "reason": "NeoBERT and Qwen3-ASR are different model additions." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT addition vs weight-sparse transformer request; different model changes." + "reason": "NeoBERT and the weight-sparse transformer request are different architectures and tasks." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "Text-model request vs audio model request; unrelated." + "reason": "NeoBERT and LFM2.5 Audio 1.5B are unrelated requests." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "CV detector addition vs text-model addition; different underlying requests." + "reason": "RF-DETR and NeoBERT are different models; no shared concrete bug/change." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are different BERT-derived model requests, not the same change." + "reason": "NeoBERT and ModernVBERT are different model additions." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture request vs Qwen3-ASR speech support; unrelated." + "reason": "MiniCPM3 architecture and Qwen3-ASR support are different model families and modalities." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "Vision-language model request vs sparse transformer request; different code paths." + "reason": "MiniCPM3 architecture and the weight-sparse transformer request are unrelated changes." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 model architecture and ModernVBERT models are separate additions." + "reason": "MiniCPM3 architecture and ModernVBERT models are different requests." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "Vision-language model request vs TTS support request; unrelated." + "reason": "MiniCPM3 architecture and Voxtral TTS support are distinct model tasks." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 detection model vs Qwen3-TTS speech model; unrelated." + "reason": "DEIMv2 and Qwen3-TTS are different model support requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT model additions vs Qwen3-ASR support; different model families and tasks." + "reason": "ModernVBERT models and Qwen3-ASR support are different model families and tasks." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT model additions vs Qwen3-TTS support; unrelated." + "reason": "ModernVBERT models and Qwen3-TTS support are unrelated." }, { "left": "issue:42503", "right": "issue:42954", "accept": false, - "reason": "ModernVBERT model request vs MiMo model request; different concrete models." + "reason": "ModernVBERT and XiaoMi MiMo are different model additions." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like RoPE support request vs Voxtral TTS support; unrelated." + "reason": "BERT-like models with RoPE and Voxtral TTS support are different feature requests." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "Both touch BERT-like text models, but one adds ModernVBERT models and the other asks for RoPE support; not the same change." + "reason": "ModernVBERT models and BERT-like RoPE support are overlapping only at a broad BERT theme, not the same change." }, { "left": "issue:42738", "right": "issue:42954", "accept": false, - "reason": "Generic RoPE request vs MiMo model request; unrelated." + "reason": "RoPE support for BERT-like models and XiaoMi MiMo are different requests." }, { "left": "issue:42738", "right": "issue:46653", "accept": false, - "reason": "Technical model request vs non-technical personal-name issue; unrelated." + "reason": "BERT-like RoPE support and Mirko Privitera are unrelated." }, { "left": "issue:42954", "right": "issue:44205", "accept": false, - "reason": "MiMo model request vs SAM3-LiteText request; different model families." + "reason": "XiaoMi MiMo and SAM3-LiteText are different model requests." }, { "left": "issue:42954", "right": "issue:43671", "accept": false, - "reason": "MiMo model request vs Qwen3-TTS support; unrelated." + "reason": "XiaoMi MiMo and Qwen3-TTS are different models and modalities." }, { "left": "issue:42954", "right": "issue:45177", "accept": false, - "reason": "MiMo model request vs DeepSeek-OCR2 request; different model types." + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are unrelated model additions." }, { "left": "issue:42954", "right": "issue:46653", "accept": false, - "reason": "Model request vs unrelated personal-name issue." + "reason": "XiaoMi MiMo and Mirko Privitera are unrelated." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are both speech-related, but they target different tasks and code paths." + "reason": "Qwen3-TTS and Qwen3-ASR are related only by naming; they are different modalities and separate support changes." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Both are TTS requests, but for different models; not the same concrete fix." + "reason": "Qwen3-TTS and Voxtral TTS are different model integrations, not the same change." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Speech TTS support vs OCR model request; unrelated." + "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model requests." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Speech model support vs weight-sparse transformer request; unrelated." + "reason": "Qwen3-TTS support and the weight-sparse transformer request are different features." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "ASR support vs audio model request; different tasks and implementations." + "reason": "Qwen3-ASR and LFM2.5 Audio 1.5B are both audio-related but are different models and different support work." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Speech ASR request vs universal perception encoder request; unrelated." + "reason": "Qwen3-ASR and EUPE are different model requests." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Speech model support vs sparse transformer request; unrelated." - }, - { - "left": "issue:43909", - "right": "issue:44205", - "accept": false, - "reason": "Audio model request vs SAM3-LiteText request; unrelated." - }, - { - "left": "issue:43909", - "right": "issue:46645", - "accept": false, - "reason": "Audio model request vs Granite SWA+sink option request; different scope." - }, - { - "left": "issue:44121", - "right": "issue:45322", - "accept": false, - "reason": "Weight-sparse transformer request vs EUPE perception encoder request; different models and changes." + "reason": "Qwen3-ASR support and the weight-sparse transformer request are unrelated." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "ASR support vs TTS support; both speech-related but not the same concrete change." + "reason": "Qwen3-ASR and Voxtral TTS are different audio-model integrations." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "Text-model request vs TTS model request; unrelated." + "reason": "NeoBERT and Voxtral TTS are unrelated model requests." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "ASR support vs OCR model request; unrelated." + "reason": "Qwen3-ASR and DeepSeek-OCR2 are different support requests." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are different vision-language/perception model requests with distinct scopes." + "reason": "EUPE and TIPSv2 are different model requests despite both being vision-language related." }, { "left": "issue:42503", "right": "issue:46653", "accept": false, - "reason": "ModernVBERT model request vs unrelated personal-name issue." + "reason": "ModernVBERT models and Mirko Privitera are unrelated." }, { "left": "issue:45027", "right": "issue:46653", "accept": false, - "reason": "TTS model request vs unrelated personal-name issue." + "reason": "Voxtral TTS support and Mirko Privitera are unrelated." } ] }, @@ -20249,8 +20193,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-38740-2", "nodes": [ @@ -20260,12 +20204,12 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "55fde3a0244cb56b3065a96f5b8353e35f60d939baf0339f49ff42de1e01f859", + "prepared_review_unit_hash": "e6efc02b755ec3068c4726697eb6e754dabb330486ee94fb32a1be05cd6df1b6", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1506, + "serialized_chars": 1505, "estimated_input_tokens": 377, "estimated_eval_tokens": 1010 }, @@ -20273,7 +20217,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1506, + "serialized_chars": 1505, "estimated_input_tokens": 377, "estimated_eval_tokens": 1010 }, @@ -20282,18 +20226,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are unrelated: one is a documentation request to add Pruna as an optimization framework, and the other is a Vision RoPE refactor. They do not describe the same bug or change, so this is not a duplicate cluster.", - "confidence": 0.99, - "canonical_issue_reason": null, + "summary": "These issues are unrelated: one is a docs request to add Pruna as an optimization framework, the other is a Vision RoPE refactor. The soft similarity signal is too weak to indicate duplication.", + "confidence": 0.98, + "canonical_issue_reason": "If a single issue must represent the cluster, issue 46443 is the more substantive engineering task; issue 38740 is a documentation-oriented request and is clearly not the same problem.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 46443 is the better representative because it describes an actual code refactor in a specific model component, whereas issue 38740 is a standalone docs enhancement.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Different topics and scopes: docs/framework listing vs. Vision RoPE code refactor; no shared target or underlying problem." + "reason": "Reject: these concerns do not describe the same underlying bug or change. One is a docs addition about Pruna, the other is a Vision RoPE refactor in a different subsystem with no shared targets or filenames." } ] }, @@ -20312,8 +20256,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41084-2", "nodes": [ @@ -20323,7 +20267,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "e1c3fe1dc7ad5b630e4fe3880ae6d47f89c4cb79f5d054a72b6f8c9e2772d6af", + "prepared_review_unit_hash": "31e8894881b851c195a93399ab05d1b027afea12e41c25f5d1afe757e28008f5", "component_budget": { "node_count": 2, "item_count": 2, @@ -20345,18 +20289,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues do not appear to be duplicates: one is a generic decoding behavior request, while the other is a model-specific compatibility/support request. The soft similarity signal is too weak to overcome the lack of shared targets, filenames, or concrete overlap in the described problem.", + "summary": "The two issues are about different topics: one requests block decoding behavior, while the other asks for DeepSeek v3.2 model support. There is no sign they describe the same underlying bug or change.", "confidence": 0.97, - "canonical_issue_reason": null, + "canonical_issue_reason": "No canonical issue: the titles and implied scopes are unrelated, with no shared targets or code-level evidence tying them together.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43037 is the more concrete, user-facing request because it names a specific target ('DeepSeek v3.2 support'), but it is not a duplicate of #41084.", + "best_issue_reason": "Neither issue is a good global best for the cluster because they do not form a duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different underlying requests: 'Set Block Decoding' describes a decoding-mode change, while 'DeepSeek v3.2 support' is about adding support for a specific model/version. They do not share a concrete code path or explicit linkage." + "reason": "These are different feature requests/problems: block decoding vs model-version support. The similarity is only superficial and does not indicate the same underlying change." } ] }, @@ -20375,8 +20319,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41306-2", "nodes": [ @@ -20386,40 +20330,40 @@ "soft_pairs": [ "issue:41306|issue:41826" ], - "prepared_review_unit_hash": "1fcf3a76168483a93596b4bf371f70093c426447a481a133f2b3ec808baad200", + "prepared_review_unit_hash": "fe7363c30fa1660cb10a7f05ea4dc3e075e2428e2135f716f0bc8b717edf3dde", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, - "estimated_input_tokens": 396, - "estimated_eval_tokens": 1048 + "serialized_chars": 1585, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, - "estimated_input_tokens": 396, - "estimated_eval_tokens": 1048 + "serialized_chars": 1585, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one requests a `from_pretrained()` flag for TiledFusedLogitsLoss/LigerFusedLinearCrossEntropy, while the other asks for TiledMLP integration to reduce memory use. They touch different model components and change requests.", - "confidence": 0.95, - "canonical_issue_reason": "Issue #41306 is the more concrete, user-facing API request and is better as the representative issue for this small cluster, but it is not a true duplicate of #41826.", + "summary": "These two issues both mention tiled/memory-saving model components, but they target different features and code paths: FLCE/LigerFusedLinearCrossEntropy flag wiring in `from_pretrained()` versus integrating TiledMLP for lower memory use. They do not look like the same underlying bug or change request.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 41826 is the broader feature request around memory-footprint reduction; 41306 is a narrower configuration/wiring request for a different component.", "canonical_pr_reason": null, - "best_issue_reason": "#41306 is narrower and better-scoped, with a specific `from_pretrained()` behavior request; #41826 is a broader integration ask. Neither covers the same underlying change.", + "best_issue_reason": "41826 is the better cluster representative because it is the more general and product-facing request about memory savings, while 41306 is a specific flag exposure problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41306", "right": "issue:41826", "accept": false, - "reason": "Different features/components: TiledFusedLogitsLoss/FLCE flag support vs TiledMLP integration. Shared 'tiled' wording is superficial and they do not describe the same bug or change." + "reason": "Both are about tiled optimization/memory reduction, but one is specifically about exposing a FLCE flag in `from_pretrained()` and the other about adding TiledMLP support. Different components and likely different implementation paths." } ] }, @@ -20438,8 +20382,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43598-2", "nodes": [ @@ -20449,7 +20393,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "a2162c90c85df7e3b02a4fe480749b3d08e97f9af4c3867e8aa3c804acbf1a25", + "prepared_review_unit_hash": "9f6090a3384c2bdef18d1f96996cb6c2a220e8ffde747d5725676de80386656b", "component_budget": { "node_count": 2, "item_count": 2, @@ -20471,18 +20415,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues both touch Trainer customization, but they describe different problems: one is about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They are related in theme but not the same underlying bug/request.", + "summary": "The two issues are both about custom trainers in `Trainer`, but they describe different problems: one is about `remove_unused_column` customizability, the other about FLOPS estimation. They are related in theme but not duplicates.", "confidence": 0.92, - "canonical_issue_reason": "No single canonical issue: the reports target different Trainer behaviors and would not be resolved by the same fix.", + "canonical_issue_reason": "Issue #43598 is the broader, more representative Trainer customizability concern; #43600 is a separate FLOPS-estimation problem.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43598 is the more concrete customization request, but it should not subsume #43600 because the FLOPS-estimation problem is separate.", + "best_issue_reason": "#43598 is the better cluster representative because it targets the core Trainer customization behavior rather than a narrower accounting edge case.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Both concern custom trainers, but one is about `remove_unused_column` customization and the other about FLOPS estimation. They do not describe the same bug or change." + "reason": "They share the general context of custom trainers, but the underlying problems differ: one concerns `remove_unused_column`, the other concerns FLOPS estimation. Not the same bug." } ] }, @@ -20501,8 +20445,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43684-2", "nodes": [ @@ -20512,12 +20456,12 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "c18592305b22a50daf50ec1e1ae78fcad5f630eb2241da6d7a2cb923d5a12204", + "prepared_review_unit_hash": "ca396277310ed4044d4b6a6c3b1d9494e027e32a1429e88c6bd5499e0e98e106", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -20525,7 +20469,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -20534,18 +20478,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These issues are not duplicates: one requests core registration for Qwen3-Omni in AutoModel/AutoModelForConditionalGeneration, while the other requests sequence-classification support for Qwen3.5 with a different config and auto head.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 43684 is the broader core registration request and is the better representative of the cluster\u2019s general auto-model support theme.", + "summary": "These are both model-registration requests, but they target different model families and different auto classes, so they do not look like the same underlying issue.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43684 is the broader and earlier request: it asks for Qwen3-Omni registration in both AutoModel and AutoModelForConditionalGeneration, which makes it the better representative of the Qwen3-Omni registration gap.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the most central and broadly scoped request; the other issue is a narrower, different-model auto-head addition.", + "best_issue_reason": "Issue 43684 is the best issue candidate because it covers a concrete registration omission and is broader in scope than the sequence-classification-only request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying changes: Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration vs Qwen3.5 AutoModelForSequenceClassification support. Different model family, config, and task head." + "reason": "Different targets: Qwen3-Omni auto-model registration versus Qwen3.5 AutoModelForSequenceClassification support. They concern different model variants and different auto-class paths, so they are not the same bug/change." } ] }, @@ -20564,8 +20508,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43835-4", "nodes": [ @@ -20582,77 +20526,80 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "1b81abe17775240fb4d73117062a1d5f4c9e101ded76359f3eafb469a77aaa55", + "prepared_review_unit_hash": "070f5b608a857c15247f2e59059837d7dbad9a05a4536d392148dbeaa54bfe0a", "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": 4267, + "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": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster looks like low-signal placeholder issues: one unrelated title ('Fett') and three near-identical but underspecified 'huggingface' issues. There isn\u2019t enough substantive detail to confidently merge any pair.", - "confidence": 0.42, - "canonical_issue_reason": "If a canonical issue must be chosen, issue 43859 is the best anchor among the vague 'huggingface' reports because it is the only one with any discussion activity; however, the report content is too thin to confirm a true duplicate cluster.", + "summary": "The duplicate cluster is most plausibly the three closed issues titled \"huggingface\"; 43835 titled \"Fett\" looks separate. The evidence here is limited to title matching and soft-similarity scores, so the duplicate call should stay conservative.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 43845 is the earliest of the three matching \"huggingface\" reports and is a reasonable representative for that duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43859 is the strongest issue to keep as the representative because it has the most activity, but the underlying problem is not described well enough to make a high-confidence choice.", + "best_issue_reason": "Issue 43845 is the best issue representative because it is the earliest item in the likely duplicate trio and matches the shared title pattern.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "The titles are different ('Fett' vs 'huggingface'), and there is no body/detail evidence showing the same underlying issue." + "reason": "Different titles; with only soft similarity and no shared targets/files, 43835 ('Fett') does not look like the same report as the \"huggingface\" issues." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "The titles are different ('Fett' vs 'huggingface'), with no explicit links or technical details to support a duplicate relationship." + "reason": "Different titles and no concrete overlap beyond soft similarity, so this does not look like the same underlying issue." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "The titles are different ('Fett' vs 'huggingface'), and the issues are too underspecified to infer the same bug." + "reason": "\"Fett\" is distinct from the repeated \"huggingface\" title, and there is no stronger evidence of a shared bug report." }, { "left": "issue:43845", "right": "issue:43846", - "accept": false, - "reason": "They share the same placeholder title, but there is no substantive description, target, or filename evidence to confirm the same underlying bug." + "accept": true, + "reason": "These have the same title and a high soft-similarity score, so they plausibly refer to the same underlying issue." }, { "left": "issue:43845", "right": "issue:43859", - "accept": false, - "reason": "The title matches, but the reports are too vague and lack technical detail to prove they describe the same issue." + "accept": true, + "reason": "These share the same title and show strong soft similarity, which is enough to treat them as likely duplicates." }, { "left": "issue:43846", "right": "issue:43859", - "accept": false, - "reason": "The title matches, but there is no concrete evidence that these are the same bug rather than similarly named placeholder reports." + "accept": true, + "reason": "Same title plus strong soft-similarity evidence makes these look like the same underlying issue report." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The clustering summary is mostly grounded, but the analyst overstates the evidence in a few places. In particular, the packet only shows matching titles and soft-similarity scores; it does not provide body text, so phrases like \u201crepeated title/body pattern\u201d and \u201cnear-identical\u201d are stronger than the evidence supports. The duplicate judgment among the three \u2018huggingface\u2019 issues is plausible, but the rationale should stay closer to the available data and be more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -20664,8 +20611,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-44371-6", "nodes": [ @@ -20690,106 +20637,106 @@ "issue:44573|issue:47036", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "e504eea0ea65dfcd9d5fb189aad886e0b92a67dc6294e45fb94596c333eaaa89", + "prepared_review_unit_hash": "0491817e9d6becd235c32fe98290b96df50e972ed17cfc49d13c3e22df575429", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7352, - "estimated_input_tokens": 1838, - "estimated_eval_tokens": 3932 + "serialized_chars": 7355, + "estimated_input_tokens": 1839, + "estimated_eval_tokens": 3934 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7352, - "estimated_input_tokens": 1838, - "estimated_eval_tokens": 3932 + "serialized_chars": 7355, + "estimated_input_tokens": 1839, + "estimated_eval_tokens": 3934 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These issues are all low-signal and appear unrelated: generic spam/doc placeholders plus one benchmark tip and one Google LLC check-out note. There is no evidence of a shared underlying bug or request, so no duplicate cluster should be formed.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "All items look like unrelated spam/placeholder issues rather than duplicates of one underlying bug or change. No PRs are present, and none of the soft-similarity pairs should be merged.", + "confidence": 0.97, + "canonical_issue_reason": "No canonical issue stands out because the items do not describe the same problem; they are mostly spammy or generic posts with unrelated titles.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 47036 is the least spam-like and the only one with any discussion, but it still does not match the others as a duplicate.", + "best_issue_reason": "No single issue is a good representative for the cluster because there is no shared technical topic to consolidate.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are generic spam titles with no concrete content or shared bug/request to unify." + "reason": "Both are generic spam placeholders, but they do not describe the same underlying issue." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Generic spam placeholders; no evidence they describe the same underlying issue." + "reason": "Both are generic spam placeholders, but they are not the same bug/change." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "A spam post and a documentation enhancement placeholder are not the same issue." + "reason": "One is spam and the other is a vague documentation suggestion; no shared underlying problem." }, { "left": "issue:44371", "right": "issue:47036", "accept": false, - "reason": "Spam placeholder vs. benchmark tip; different topics and no shared problem." + "reason": "Unrelated spam placeholder versus community tip; not a duplicate report." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder and a Google LLC check-out note are unrelated." + "reason": "Different unrelated spam content; no common bug or change." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Documentation enhancement suggestion and spam placeholder are not the same underlying request." + "reason": "A vague documentation idea is not the same as a spam placeholder issue." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "Documentation enhancement placeholder does not match a spam issue." + "reason": "These do not describe the same concrete problem or change." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are spam titles, but there is no concrete evidence of a single shared bug/change." + "reason": "Both are spam placeholders, but they are distinct unrelated posts rather than one report." }, { "left": "issue:44572", "right": "issue:47036", "accept": false, - "reason": "Spam placeholder versus benchmark tip; no common code-path or request." + "reason": "Spam placeholder versus benchmark tip; no shared underlying issue." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder and Google LLC check-out note are clearly different topics." + "reason": "Different unrelated spam content, not a duplicate." }, { "left": "issue:44573", "right": "issue:47036", "accept": false, - "reason": "Spam placeholder and benchmark tip are unrelated." + "reason": "These are unrelated posts and do not target the same problem." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder and Google LLC check-out note do not look like duplicates." + "reason": "Different content and intent; not the same issue." } ] }, @@ -20808,8 +20755,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45145-2", "nodes": [ @@ -20819,7 +20766,7 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "cd7c5b9e4e2f796170a8039fb075d69ab14266928bdab3df7e07e0375f0f2523", + "prepared_review_unit_hash": "1fa3478aa249d6401b7780a4d0bccd30a0030a90cc6ad66e4116c3eb9c341985", "component_budget": { "node_count": 2, "item_count": 2, @@ -20841,24 +20788,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues are likely duplicates or very closely related: they share the exact title and have moderate textual similarity, but there isn\u2019t enough evidence to call them near-exact duplicates. #45145 is the better representative because it has the inbound reference, while #45151 has no linkage signal.", + "summary": "The two issues are likely duplicates: they share the exact title, were opened within hours of each other, have very similar body lengths, and the pair has a soft_similarity signal. #45145 is the best representative because it is earlier and has the only inbound reference.", "confidence": 0.79, - "canonical_issue_reason": "#45145 is the better canonical issue because it has the same title as #45151 and at least one inbound reference, making it the slightly more established thread.", + "canonical_issue_reason": "Choose #45145 as canonical because it is the earlier filing and has slightly more link signal (one inbound reference), making it the better representative for this likely duplicate report.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best issue to keep as the cluster representative due to the identical title and stronger linkage history from the inbound reference.", + "best_issue_reason": "#45145 is the best issue representative because it was opened first and has the stronger external signal among the two; #45151 appears to be a later duplicate of the same report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": true, - "reason": "The issues share an identical title and have substantial textual similarity, with no contrary signals like different targets or filenames. The evidence is not strong enough to call them exact duplicates, but it is sufficient to treat them as the same underlying issue for triage." + "reason": "Same title, close creation times, similar body lengths, and a soft_similarity signal make this likely a duplicate issue pair, even though the evidence is not strong enough to be definitive." } ] }, "evaluator_result": { "accept": false, - "feedback": "The packet supports a likely duplicate/closely related pair, but the analyst overstates certainty. The only similarity evidence is same title plus a soft_similarity signal with Jaccard 0.63 and deterministic_accept=false; that is not enough to call them \"near-exact duplicates\" or say there is \"no evidence\" of different bugs. The canonical choice based on the single inbound reference is grounded, but the summary and soft-edge rationale should be more conservative." + "feedback": "The conclusion is plausible, but the analyst overstates the evidence. The packet only directly supports same title, close creation times, similar body length, and a soft_similarity signal; it does not show the bodies are actually 'strongly similar' or that there is 'no distinguishing metadata.' The 0.98 confidence is also too strong for this level of evidence. The soft-edge accept=true is a bit too assertive given deterministic_accept=false and only moderate similarity signal." }, "evaluator_used": true, "retried": true, @@ -20928,8 +20875,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45762-2", "nodes": [ @@ -20939,12 +20886,12 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "81a99dbb81a76471886eb4ba3b4f2e048aab925f7e5c43869c30c54f25cc6d81", + "prepared_review_unit_hash": "ffe7623028f1e98b327a55a36c73f25b2c3ed9ba971e0774de75e730655ce96e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1475, + "serialized_chars": 1476, "estimated_input_tokens": 369, "estimated_eval_tokens": 994 }, @@ -20952,7 +20899,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1475, + "serialized_chars": 1476, "estimated_input_tokens": 369, "estimated_eval_tokens": 994 }, @@ -20961,18 +20908,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues appear unrelated: both are bare name/date titles with no problem description, no shared references, and no clear underlying bug or feature change to deduplicate.", + "summary": "These two closed issues look like unrelated spam/personal-name posts, not duplicates of the same bug report. There is no clear canonical issue in this cluster.", "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Neither issue is a meaningful bug report or a better representative of a shared problem; both appear to be standalone spam/irrelevant entries.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "The titles reference different people and different dates, with no issue body or shared context indicating the same bug or request. This does not look like a duplicate pair." + "reason": "The titles are different personal names with different dates and contain no shared bug context, target, or description indicating the same underlying issue." } ] }, @@ -20991,8 +20938,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45764-2", "nodes": [ @@ -21002,7 +20949,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "88e4be169201e641991ae6d8ced2caa747a4d90525d209cebd2ed2e15bba49ec", + "prepared_review_unit_hash": "137225402133f273d8829f6743d80b0fc649d8160a718b1254a8e4b09f8b658f", "component_budget": { "node_count": 2, "item_count": 2, @@ -21024,18 +20971,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two closed issues with only personal-name-and-date titles; they do not describe a product bug, feature, or the same underlying problem. This looks like unrelated/spam-like content rather than a duplicate cluster.", + "summary": "The two closed issues are unrelated, name/date-style posts with no technical content or shared problem statement, so they should not be clustered as duplicates.", "confidence": 0.98, - "canonical_issue_reason": "No clear canonical issue: the two issues are not about the same topic and contain no actionable bug report content.", - "canonical_pr_reason": "No PRs are present in this cluster.", - "best_issue_reason": "Neither issue is suitable as a canonical representative because they are unrelated and non-technical.", - "best_pr_reason": "No PRs are present in this cluster.", + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "Different names and dates, with no shared bug description, code path, or explicit linkage; not the same underlying issue." + "reason": "Different named entries with different dates and no body/context indicating the same underlying issue; this does not look like a duplicate." } ] }, @@ -21054,8 +21001,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45810-2", "nodes": [ @@ -21065,40 +21012,40 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "772036ee9baf571ebc9129ee0a6f7f63a2cf0592bced4e587ebe9dfabf55f79c", + "prepared_review_unit_hash": "27b884051499e6204dc95cf09f9a35183a96f22dbafece899c6fba8d11845b72", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1542, - "estimated_input_tokens": 386, - "estimated_eval_tokens": 1028 + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1542, - "estimated_input_tokens": 386, - "estimated_eval_tokens": 1028 + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one requests a new Qwen3_5 token-classification head for value modeling, while the other asks for FP8 training support. They concern different features and different code paths.", + "summary": "These two issues are not duplicates: one requests a new Qwen3_5 token-classification model class, while the other requests FP8 training support. They touch different features and code paths.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "issue:46295 is the broader, more generally scoped request and has more inbound discussion, so it is the better representative issue in this set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45810 is the more concrete, narrowly scoped request, but it is not a valid canonical representative for this cluster because the two issues are unrelated.", + "best_issue_reason": "issue:46295 is the most central standalone issue here because it appears broader and more active than the model-addition request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Reject: the titles and descriptions point to different features (model architecture/head support vs FP8 training support) with no shared underlying bug or change." + "reason": "Different underlying requests: adding a token-classification model class vs enabling FP8 training support. They do not describe the same bug or change." } ] }, @@ -21117,8 +21064,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45869-5", "nodes": [ @@ -21138,88 +21085,88 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "6a58c964f620abcbf1ffa6c9f179103c4877c525068e79c24fa19e1f107aebe6", + "prepared_review_unit_hash": "64806193afc2b183845d500e1f23ea840e9531fd4d2a51e692e972b92dae658c", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, - "estimated_input_tokens": 1432, - "estimated_eval_tokens": 3120 + "serialized_chars": 5724, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, - "estimated_input_tokens": 1432, - "estimated_eval_tokens": 3120 + "serialized_chars": 5724, + "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 contains two separate exact-duplicate issue pairs: one about a DataForge financial time-series example, and one about a DataForge financial time-series + scraping service request. The dataset-bundle promo issue is related by theme only, not the same underlying request.", - "confidence": 0.96, - "canonical_issue_reason": "issue:45872 is the clearest representative of the duplicate pair for the 'financial time series + scraping service' request: it is specific, actionable, and has an exact duplicate (issue:45873).", + "summary": "There are two clear duplicate issue pairs: 45870/45871 and 45872/45873. The dataset-bundle issue (45869) is only loosely related to the others and should not be merged with them.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 45870 is a good canonical representative because it is the earliest exact duplicate pair for the repeated \u201cExample: Financial time series prediction with DataForge\u201d report and is cleaner to retain as the primary thread.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45872 is the strongest overall representative in this cluster because it is a concrete, duplicate-backed feature request rather than a broad promo/example post.", + "best_issue_reason": "45870 is the best single issue to keep as the representative item among the near-identical reports; it has an exact duplicate and the title is the clearest standalone example.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Different artifacts: one is a dataset promotion, the other is an example request. Same financial-data theme, but not the same underlying issue." + "reason": "Same general finance/dataset theme, but this looks like a dataset offering versus an example post; not the same underlying issue." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Different artifacts: dataset bundle promo vs example request. Similar topic, but no evidence they describe the same change or bug." + "reason": "Related topic, but the dataset-bundle post is not the same concrete report as the example issue." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Exact duplicate: same title and no distinguishing metadata suggesting different requests." + "reason": "Exact title match and same body shape; these are the same issue." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Both mention financial time series and DataForge, but one is an example and the other is a scraping-service feature request. Not the same underlying issue." + "reason": "Both mention DataForge and finance, but one is an example and the other is a scraping-service post; different underlying requests." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Same as issue:45870 vs issue:45872: related topic, but different request scope and no evidence of a shared concrete change." + "reason": "Same as 45870/45872: only broadly similar theme, not the same concrete issue." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "The titles point to different requests: example notebook/article versus scraping-service feature. Not a duplicate." + "reason": "Exact duplicate of 45870, but not the same underlying issue as the scraping-service request." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Different requests despite the shared financial time-series theme; no indication they should be merged." + "reason": "Same reason as 45871/45872; different request despite shared finance/DataForge wording." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Exact duplicate: same title and same body length/metadata pattern, indicating the same underlying request." + "reason": "Exact title match and same body shape; these are the same issue." } ] }, "evaluator_result": { "accept": true, - "feedback": "The cluster summary is grounded in the packet: it correctly separates the dataset promo from the two duplicate issue pairs, and the soft-edge verdicts are conservative overall. The duplicate-pair calls for 45870/45871 and 45872/45873 are supported by identical titles plus matching body lengths and jaccard 1.0, while the cross-theme pairs are reasonably rejected." + "feedback": "The summary is well grounded in the packet: the identical-title, jaccard-1.0 pairs 45870/45871 and 45872/45873 are reasonably treated as duplicate clusters, while 45869 is only loosely related and kept separate. The soft-edge verdicts are conservative and align with the available evidence." }, "evaluator_used": true, "retried": false, @@ -21235,8 +21182,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46439-2", "nodes": [ @@ -21246,7 +21193,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "8e1f3dc1157e617811f01ee6724a600f2b29d7a997631efee0c85103178ea779", + "prepared_review_unit_hash": "79b1e5f076b6b55be1de1645a7e0b60bc71d0dc1c6aaa634d9bf2741d5bdab4e", "component_budget": { "node_count": 2, "item_count": 2, @@ -21268,18 +21215,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are closely related to static-cache behavior in linear-attention models, but they describe different concrete failures: one is a shape corruption/runtime error during early initialization, the other is a `KeyError: 'linear_attention'` on Qwen3.5. Not enough evidence they are the same underlying bug.", - "confidence": 0.82, - "canonical_issue_reason": "Issue #46439 is the better canonical issue because it describes the more specific code-path bug in static cache `early_initialization` and gives a clearer failure mode to anchor investigation.", + "summary": "Both issues concern `cache_implementation=\"static\"` and `linear_attention`, but they describe different failure modes: one is a shape-corruption `RuntimeError`, the other a `KeyError` in Qwen3.5. Related, but not clearly the same underlying bug from the available evidence.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it describes the broader root-cause class (`early_initialization` breaking linear-attention layers) rather than a single model-specific crash symptom.", "canonical_pr_reason": null, - "best_issue_reason": "#46439 is the strongest representative since it focuses on the likely root-cause path rather than a model-specific symptom.", + "best_issue_reason": "Issue 46439 is the best issue candidate overall: it is more general, points at the likely underlying static-cache initialization problem, and can subsume narrower reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "Both mention static cache and linear attention, but one reports corrupted shapes from `early_initialization` while the other reports a `KeyError: 'linear_attention'` on Qwen3.5. These look like related symptoms, not clearly the same bug or fix." + "reason": "Same area, but the reported bugs differ concretely (`RuntimeError` shape corruption vs `KeyError: 'linear_attention'`), so there is not enough evidence they are the same underlying defect." } ] }, @@ -21298,8 +21245,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46551-2", "nodes": [ @@ -21309,7 +21256,7 @@ "soft_pairs": [ "issue:46551|issue:46552" ], - "prepared_review_unit_hash": "e64d0f51bde8757a7e689e2b010aeb76910de553f7b610d051612673b909c153", + "prepared_review_unit_hash": "66c3e2d38afcea752b232ce89d261d55dac4eac22e82b348091e36dd403f0ce9", "component_budget": { "node_count": 2, "item_count": 2, @@ -21331,24 +21278,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are duplicate reports of the same Wav2Vec2CTCTokenizer v5 regression; the open issue is the better canonical tracker.", + "summary": "The two issues are exact duplicates: same title, same body, same bug in Wav2Vec2CTCTokenizer v5. Issue #46552 is the better canonical issue because it is open and already has inbound references, while #46551 is closed.", "confidence": 0.99, - "canonical_issue_reason": "issue:46552 is the active open report and already has inbound references, so it is the better single tracker for the bug.", + "canonical_issue_reason": "#46552 is the stronger canonical issue: it is open, has inbound references, and matches the same exact report as #46551.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46552 is the best issue representative because it is open, current, and appears to be the same report as the closed duplicate.", + "best_issue_reason": "#46552 best represents the cluster because it is the active report for the identical tokenizer special-token leak bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46551", "right": "issue:46552", "accept": true, - "reason": "Same title, same described tokenizer bug, and no distinguishing details; these are duplicate reports of the same underlying issue." + "reason": "Exact duplicate issue reports: identical title, identical body, and the same underlying Wav2Vec2CTCTokenizer v5 special-token bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: the two issues have identical titles, identical body length, and a perfect soft-similarity match, with one closed and one open. Treating the open issue as the canonical tracker and the closed one as its duplicate is conservative and supported by the evidence." + "feedback": "The summary is well-grounded in the packet: both issues have the same title, the same body length, and a Jaccard similarity of 1.0, so treating them as duplicates is reasonable. The canonical choice of #46552 is also supported by the packet because it is open and has inbound references, while #46551 is closed. No mergeability issues apply since this is an issue pair, not a PR pair." }, "evaluator_used": true, "retried": false, @@ -21364,8 +21311,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46657-2", "nodes": [ @@ -21375,40 +21322,40 @@ "soft_pairs": [ "issue:46657|issue:46692" ], - "prepared_review_unit_hash": "4c39fde43788f59b3257695a16d79e580a73d6da0a7cb771a70b6d123e9ae5c3", + "prepared_review_unit_hash": "fde121ff9bf08ef0ccbb105bb260e155feedd88c5b7f3405d3836f6251ddfa9c", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1569, - "estimated_input_tokens": 393, - "estimated_eval_tokens": 1042 + "serialized_chars": 1568, + "estimated_input_tokens": 392, + "estimated_eval_tokens": 1040 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1569, - "estimated_input_tokens": 393, - "estimated_eval_tokens": 1042 + "serialized_chars": 1568, + "estimated_input_tokens": 392, + "estimated_eval_tokens": 1040 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not the same bug: one reports a tensor size mismatch during object-detection training, while the other reports out-of-memory after several epochs in RT-DETR v2 training. The soft similarity is too broad to treat them as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No clear canonical issue for a duplicate cluster: the two issues describe different failure modes and different training symptoms.", + "summary": "These two issues are in the same broad training/object-detection area, but they describe different problems: a tensor shape mismatch versus an out-of-memory condition after several epochs. They should not be deduplicated.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 46657 is the more specific, concrete report and best anchor for the cluster; 46692 describes a different failure mode (OOM during RT-DETR v2 training) rather than the same underlying bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #46657 is the cleaner representative bug report because it describes a concrete tensor-shape mismatch error, whereas #46692 is a separate OOM symptom.", + "best_issue_reason": "46657 is the better representative issue because it names a precise runtime error ('size of tensors mismatch'). 46692 appears to be a separate resource/memory issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46657", "right": "issue:46692", "accept": false, - "reason": "Different underlying problems: tensor size mismatch versus out-of-memory after some epochs. They do not appear to be the same bug." + "reason": "Reject: the titles point to different underlying bugs and symptoms. A tensor-size mismatch during object-detection training is not the same as memory exhaustion after some epochs in RT-DETR v2, even though both occur in training." } ] }, @@ -21427,8 +21374,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46822-2", "nodes": [ @@ -21438,7 +21385,7 @@ "soft_pairs": [ "issue:46822|issue:46823" ], - "prepared_review_unit_hash": "7accf896c99dd02025071c6611a1298f2d74202bb50a54ddfb3b7a0bce5d260f", + "prepared_review_unit_hash": "ab1dfb8e4df510766853854e845d4e332838cdd4bfc22e1323e896fc929d2408", "component_budget": { "node_count": 2, "item_count": 2, @@ -21460,18 +21407,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues describe different failure modes in different loading paths: ZeRO-3 zero.Init partitioning for a composite multimodal submodule vs. 4-bit auto device mapping incorrectly offloading to CPU. They are not duplicates.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "These are two different loading bugs in different code paths: ZeRO-3 zero.Init partitioning/OOM vs 4-bit device_map/max_memory CPU dispatch. They should not be treated as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue; the two issues describe distinct failure modes in separate subsystems (ZeRO-3 partitioning vs 4-bit auto-dispatch behavior).", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46823 is the better representative of the pair only because it is phrased as a self-contained, broadly reproducible loader/device-map bug; however, it is still unrelated to issue 46822.", + "best_issue_reason": "Neither issue is a good cluster representative for the other because they are unrelated bugs and not mergeable into a single duplicate target.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46822", "right": "issue:46823", "accept": false, - "reason": "Different subsystems and root causes: DeepSpeed ZeRO-3 zero.Init partitioning of a composite model submodule vs. 4-bit device_map/max_memory dispatch to CPU. They do not look like the same underlying bug." + "reason": "Different underlying bugs: one is ZeRO-3 zero.Init failing to partition a composite submodule, the other is 4-bit loading with device_map/max_memory sending modules to CPU incorrectly." } ] }, @@ -21490,8 +21437,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46906-9", "nodes": [ @@ -21523,231 +21470,207 @@ "issue:46969|issue:46998", "issue:46985|issue:47037", "issue:46985|issue:47049", - "issue:46998|issue:47037", "issue:46998|issue:47053", + "issue:46998|issue:47037", "issue:46998|issue:47049", "issue:46985|issue:46998", - "issue:47037|issue:47053", "issue:47037|issue:47049", + "issue:47037|issue:47053", "issue:47049|issue:47053", "issue:46969|issue:47049", "issue:46985|issue:47053", "issue:46969|issue:47053" ], - "prepared_review_unit_hash": "a44dcee99a6d81647cd66bac5b7a94b7b567f7ca8d87434f34ca3d07aacbea5d", + "prepared_review_unit_hash": "6284bb4510c0fb40caacb43971c8f48d3f4a2e31deac9bd6cecca40555e357d1", "component_budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 27, - "serialized_chars": 14781, - "estimated_input_tokens": 3696, - "estimated_eval_tokens": 7648 + "serialized_chars": 14785, + "estimated_input_tokens": 3697, + "estimated_eval_tokens": 7650 }, "budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 27, - "serialized_chars": 14781, - "estimated_input_tokens": 3696, - "estimated_eval_tokens": 7648 + "serialized_chars": 14785, + "estimated_input_tokens": 3697, + "estimated_eval_tokens": 7650 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are recurring daily [serge] integration failure triage issues, not duplicates of the same underlying report. They share a template and topic, but each issue is date-specific and appears to track a different day\u2019s failures.", - "confidence": 0.95, - "canonical_issue_reason": "Issue 46906 is the earliest in the series and has the strongest reference signal, making it a reasonable representative if one must be picked from this recurring triage set.", + "summary": "These look like a dated series of daily triage issues for the same area, but the available metadata does not show a shared underlying bug or fix. I would not merge them as duplicates on title/template similarity alone.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 46906 is the earliest in the series and has the strongest inbound reference count, so it is the best anchor if the cluster is being represented by a single issue.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 47053 is the most recent and most active (highest discussion activity among the later entries), so it is the best single issue to surface as the latest triage thread.", + "best_issue_reason": "Among the issues, 46906 is the most representative hub: earliest created and most-linked, with no evidence that another issue is a better canonical representative.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46906", "right": "issue:46940", "accept": false, - "reason": "Same triage template, but the titles are for different dates; these are separate daily reports, not one bug." + "reason": "Same triage template and adjacent dates, but no evidence they track the same underlying failure; likely separate daily reports." }, { "left": "issue:46906", "right": "issue:46948", "accept": false, - "reason": "Date-stamped triage threads for different days; similarity reflects the template, not the same underlying incident." + "reason": "Strong textual similarity, but these read as distinct day-specific triage entries rather than one duplicate defect." }, { "left": "issue:46906", "right": "issue:46969", "accept": false, - "reason": "Recurring daily integration-failure triage, but each issue tracks a distinct day\u2019s failures." + "reason": "Similar series title only; no body or diff evidence that the underlying incident is the same." }, { "left": "issue:46906", "right": "issue:46985", "accept": false, - "reason": "Different triage date in the title indicates a separate report, not a duplicate." + "reason": "Date-stamped triage posts are not enough to establish duplicate status; likely separate reports." }, { "left": "issue:46906", "right": "issue:47037", "accept": false, - "reason": "These are sequential daily triage issues rather than duplicate reports of one problem." + "reason": "Both are triage issues, but the metadata only supports a recurring template, not one shared bug." }, { "left": "issue:46940", "right": "issue:46948", "accept": false, - "reason": "Same pattern and subsystem, but different date-specific triage issues." + "reason": "Adjacent triage posts with the same template, but no concrete evidence of the same failure mode." }, { "left": "issue:46940", "right": "issue:46969", "accept": false, - "reason": "The titles show separate daily triage threads, so they should not be merged as duplicates." + "reason": "Looks like another daily triage entry in the series, not a duplicate of a specific underlying issue." }, { "left": "issue:46940", "right": "issue:46985", "accept": false, - "reason": "Different days\u2019 integration failure triage; similarity is only from the recurring template." + "reason": "High similarity is explained by the repeated triage format; no shared bug details are available." }, { "left": "issue:46940", "right": "issue:47049", "accept": false, - "reason": "Not the same underlying bug\u2014just the same daily triage format across dates." + "reason": "Same broad subsystem and template, but not enough to say these are the same incident." }, { "left": "issue:46948", "right": "issue:46969", "accept": false, - "reason": "Date-stamped reports for different days; no evidence they describe the same incident." + "reason": "Similar dated triage entries, but nothing indicates the same root cause or resolution target." }, { "left": "issue:46948", "right": "issue:46985", "accept": false, - "reason": "These are separate triage entries, not duplicate bug reports." + "reason": "These appear to be separate daily reports rather than duplicate issues." }, { "left": "issue:46948", "right": "issue:46998", "accept": false, - "reason": "Same recurring triage series, but each issue corresponds to a different day." + "reason": "Template similarity only; no evidence of one underlying bug being tracked twice." }, { "left": "issue:46969", "right": "issue:46985", "accept": false, - "reason": "The pair is only similar because both are daily triage threads; they are distinct reports." + "reason": "Same triage-series pattern, but not enough signal to merge as duplicates." }, { "left": "issue:46969", "right": "issue:46998", "accept": false, - "reason": "Different dates and likely different failure snapshots; do not treat as duplicates." + "reason": "The titles match a recurring report format; no concrete shared failure is visible." }, { "left": "issue:46969", "right": "issue:47037", "accept": false, - "reason": "Same template, different day; no concrete shared incident to merge." + "reason": "Likely separate daily triage items; similarity is too generic to prove duplication." }, { "left": "issue:46969", "right": "issue:47049", "accept": false, - "reason": "Recurring triage issues for separate dates, not one underlying bug." + "reason": "No evidence beyond the repeated triage naming convention that these describe the same bug." }, { "left": "issue:46969", "right": "issue:47053", "accept": false, - "reason": "Distinct daily triage thread with a later date; not the same report." + "reason": "Same series, but the available metadata does not establish a single underlying defect." }, { "left": "issue:46985", "right": "issue:46998", "accept": false, - "reason": "Different daily integration-failure triage entries." + "reason": "Two consecutive triage reports; similar wording is expected and does not imply duplicate status." }, { "left": "issue:46985", "right": "issue:47037", "accept": false, - "reason": "The titles indicate different dates, so these are separate reports." + "reason": "No body/diff evidence of the same code-path problem, only the shared triage template." }, { "left": "issue:46985", "right": "issue:47049", "accept": false, - "reason": "No evidence of a shared concrete failure; only the shared triage template." + "reason": "These look like separate daily issue reports, not one issue reopened under a new date." }, { "left": "issue:46985", "right": "issue:47053", "accept": false, - "reason": "Later date-specific triage issue, not a duplicate of the earlier one." + "reason": "Insufficient evidence of the same underlying failure; the match is only in the recurring title format." }, { "left": "issue:46998", "right": "issue:47037", "accept": false, - "reason": "Same recurring series, but not the same underlying bug or incident." - }, - { - "left": "issue:46998", - "right": "issue:47053", - "accept": false, - "reason": "Different day\u2019s triage report; similarity is superficial." + "reason": "Recurring integration-triage formatting is not enough to treat these as duplicates." }, { "left": "issue:46998", "right": "issue:47049", "accept": false, - "reason": "These track separate daily failures, so they should remain distinct." - }, - { - "left": "issue:46985", - "right": "issue:46998", - "accept": false, - "reason": "Same issue family, but the date-stamped titles show separate triage runs." + "reason": "The metadata suggests a series of reports, not one repeated bug report." }, { - "left": "issue:47037", + "left": "issue:46998", "right": "issue:47053", "accept": false, - "reason": "Adjacent daily triage issues, not the same concrete code-path problem." + "reason": "Same template, different day; no concrete shared defect evidence." }, { "left": "issue:47037", "right": "issue:47049", "accept": false, - "reason": "Distinct date-specific reports; merge would conflate separate incidents." - }, - { - "left": "issue:47049", - "right": "issue:47053", - "accept": false, - "reason": "Different daily triage threads with no evidence of a single underlying defect." - }, - { - "left": "issue:46969", - "right": "issue:47049", - "accept": false, - "reason": "The shared wording comes from a repeating triage template, not duplicate content." + "reason": "Similar titles alone do not show the same underlying failure or fix." }, { - "left": "issue:46985", + "left": "issue:47037", "right": "issue:47053", "accept": false, - "reason": "These are separate daily integration-failure triage issues, not one bug report." + "reason": "Likely separate triage entries in a daily sequence rather than duplicates." }, { - "left": "issue:46969", + "left": "issue:47049", "right": "issue:47053", "accept": false, - "reason": "Different dates and separate triage context mean they are not duplicates." + "reason": "Adjacent date-stamped triage reports; no indication they are the same issue." } ] }, @@ -21766,8 +21689,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46983-2", "nodes": [ @@ -21777,7 +21700,7 @@ "soft_pairs": [ "issue:46983|issue:46984" ], - "prepared_review_unit_hash": "8d482e77278e2cb7ea90315409fc14dbe0409f117ac29d3378ab6c2fcf31787b", + "prepared_review_unit_hash": "f78da9fb09e1e776166274b0a1415aa095bfdd39e6672e232ac37d3ae2ee14d4", "component_budget": { "node_count": 2, "item_count": 2, @@ -21799,27 +21722,27 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are near-exact duplicate issues requesting RWKV7 model support; the open issue is the better canonical tracker.", - "confidence": 0.99, - "canonical_issue_reason": "Issue #46984 is open and can continue to track the request; #46983 is a closed duplicate with the same title and apparent scope.", + "summary": "These look like likely duplicates of the same RWKV7 support request: identical titles, very close creation times, and a strong similarity signal. The evidence is limited because the packet does not expose body text or linking metadata, so the conclusion should be treated as provisional rather than certain.", + "confidence": 0.74, + "canonical_issue_reason": "#46984 is the better canonical issue thread because it is open and appears to be the surviving discussion for the same request, but this is based mainly on the identical title, near-simultaneous creation time, and soft similarity signal since the body content is not visible.", "canonical_pr_reason": null, - "best_issue_reason": "#46984 is the best issue because it is still open and serves as the active discussion/triage point for the same RWKV7 support request.", + "best_issue_reason": "#46984 is the best issue to keep because it is open and likely represents the active thread for this same RWKV7 support request; the available evidence is strong enough for a duplicate triage, but not enough to claim full certainty.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46983", "right": "issue:46984", "accept": true, - "reason": "Same title, same requested feature, and no evidence of distinct scope; these look like duplicate reports of the same underlying request." + "reason": "Same title, created minutes apart, and a strong similarity score make them likely the same underlying request. There is no body text or cross-link evidence in the packet, so this is a cautious accept rather than a definitive one." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The conclusion is grounded in the packet: both issues share the exact title, the similarity signal is high, and one is already closed while the other remains open, so using the open issue as the canonical tracker is conservative. The soft-edge duplicate verdict is also appropriately cautious given the available evidence." + "accept": false, + "feedback": "The duplicate/canonical choice is plausible, but the writeup is a bit overconfident for the evidence available. The packet only shows same title, close timestamps, and a soft similarity signal; it does not expose body content, and there is a meaningful metadata difference (closed vs open, plus different discussion activity). Please phrase the conclusion more conservatively and lower the confidence unless you can cite stronger evidence from the bodies or links." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -21832,8 +21755,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-47007-2", "nodes": [ @@ -21843,12 +21766,12 @@ "soft_pairs": [ "issue:47007|issue:47015" ], - "prepared_review_unit_hash": "514748d2c3c1e44147a12aea5532dd44c336649fa001866165a53494804bb9ee", + "prepared_review_unit_hash": "b7bfdf0af3640fe8affd04e43152543eb066c5432b84ab2ba4a10b3fc46e661d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1626, + "serialized_chars": 1625, "estimated_input_tokens": 407, "estimated_eval_tokens": 1070 }, @@ -21856,7 +21779,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1626, + "serialized_chars": 1625, "estimated_input_tokens": 407, "estimated_eval_tokens": 1070 }, @@ -21865,18 +21788,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are thematically close, but they read like different high-level writeups rather than the same underlying request or bug. The overlap is broad AI/cybersecurity language, not a concrete shared change target.", - "confidence": 0.87, - "canonical_issue_reason": null, + "summary": "Both items are broadly about AI-assisted cybersecurity vulnerability discovery, but they read like distinct writeups rather than the same underlying issue. I would not merge them as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "Issue 47015 is the broader, more general formulation of the topic, so it is the better representative if a single issue must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "issue:47015 is the broader and clearer representative of the shared theme; issue:47007 is more of a lessons-learned framing and looks less canonical for the cluster.", + "best_issue_reason": "Issue 47015 is the most general and concise description of the shared theme, but the pair does not look like a true duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47007", "right": "issue:47015", "accept": false, - "reason": "Similar subject matter, but the titles point to different angles: one is lessons learned from advanced LLMs, the other is a novel transformer application. Not enough evidence they are the same underlying issue." + "reason": "The titles overlap on AI-driven vulnerability discovery, but they describe different angles ('lessons from using advanced LLMs' vs 'novel application of transformers') and there is no evidence they refer to the same underlying report or change." } ] }, @@ -21898,7 +21821,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-47", + "cluster_id": "cluster-40546-48", "nodes": [ "pull_request:43451", "pull_request:45694", @@ -21943,7 +21866,8 @@ "pull_request:47062", "pull_request:47069", "pull_request:47077", - "pull_request:47078" + "pull_request:47078", + "pull_request:47079" ], "soft_pairs": [ "pull_request:46903|pull_request:46958", @@ -21960,8 +21884,8 @@ "pull_request:46927|pull_request:47055", "pull_request:46908|pull_request:47006", "pull_request:46908|pull_request:46928", - "pull_request:46927|pull_request:46986", "pull_request:46907|pull_request:46926", + "pull_request:46927|pull_request:46986", "pull_request:46986|pull_request:47058", "pull_request:46926|pull_request:46928", "pull_request:46928|pull_request:47057", @@ -21987,18 +21911,18 @@ "pull_request:46949|pull_request:46968", "pull_request:47077|pull_request:47078", "pull_request:46266|pull_request:46957", - "pull_request:47062|pull_request:47077", "pull_request:47019|pull_request:47025", + "pull_request:47062|pull_request:47077", "pull_request:47034|pull_request:47035", - "pull_request:46196|pull_request:46995", "pull_request:47019|pull_request:47069", + "pull_request:46196|pull_request:46995", "pull_request:46957|pull_request:46965", "pull_request:46266|pull_request:46707", "pull_request:46707|pull_request:46961", "pull_request:46968|pull_request:47019", "pull_request:46707|pull_request:46968", - "pull_request:46707|pull_request:46957", "pull_request:46912|pull_request:47025", + "pull_request:46707|pull_request:46957", "pull_request:46707|pull_request:47019", "pull_request:46968|pull_request:47062", "pull_request:46961|pull_request:47019", @@ -22012,246 +21936,390 @@ "pull_request:46266|pull_request:47034", "pull_request:46707|pull_request:47025", "pull_request:45694|pull_request:47069", - "pull_request:46836|pull_request:46961", + "pull_request:43451|pull_request:47079", "pull_request:46836|pull_request:46965", + "pull_request:46836|pull_request:46961", "pull_request:46995|pull_request:47003", "pull_request:47035|pull_request:47069", - "pull_request:46957|pull_request:46961", - "pull_request:46961|pull_request:47035" + "pull_request:46957|pull_request:46961" ], - "prepared_review_unit_hash": "d9aebe43095e170f3caca08b38cc35a3813154dca13975b6be037279b996a97c", + "prepared_review_unit_hash": "237bb7d60a04a3dba0e4cf1fdf7b2a43ea68274f8424b347e3d846e92fcc7ec0", "component_budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 156, - "serialized_chars": 173692, - "estimated_input_tokens": 43423, - "estimated_eval_tokens": 87102 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 }, "budget": { - "node_count": 44, - "item_count": 44, + "node_count": 45, + "item_count": 45, "soft_pair_count": 72, - "serialized_chars": 124090, - "estimated_input_tokens": 31023, - "estimated_eval_tokens": 62302 + "serialized_chars": 126069, + "estimated_input_tokens": 31518, + "estimated_eval_tokens": 63292 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous cluster of unrelated PRs (docs, model additions, infra, and many model-specific test/fix patches). Only a few soft edges look like true duplicate/subset-superset fixes for the same concrete model bug; the rest should stay separate.", - "confidence": 0.84, + "summary": "This cluster is not a duplicate set: it mixes unrelated model additions, bug fixes, tests, docs, and automation changes. None of the soft edges look like the same underlying fix or change, so all are rejected.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR cleanly represents the whole cluster because it mixes unrelated changes; if a representative is needed, PR 46707 is the broadest code-centric anchor (distributed/generation/trainer infrastructure).", + "canonical_pr_reason": "No canonical PR: the items are mostly unrelated fixes, feature additions, and docs/test tweaks rather than one duplicate change.", "best_issue_reason": null, - "best_pr_reason": "PR 46707 is the strongest global representative by scope and centrality, but it is still not a duplicate anchor for the cluster as a whole.", + "best_pr_reason": "No single PR cleanly represents the cluster as a duplicate group; the highest-scope changes still overlap only loosely with a few neighbors.", "soft_edge_verdicts": [ { "left": "pull_request:46903", "right": "pull_request:46958", - "accept": true, - "reason": "Same CohereASR training-loss double-shift fix; 46903 is broader, but the core cohere_asr change is the same." + "accept": false, + "reason": "Same model family, but different loss implementations and different bug paths: shift-label handling vs manual cross-entropy." }, { "left": "pull_request:46950", "right": "pull_request:46951", "accept": false, - "reason": "Different problems: Mamba2 OOM vs PEFT adapter loading behavior." + "reason": "Mamba2 OOM/kernel behavior is unrelated to PEFT checkpoint loading and mismatched-size handling." }, { "left": "pull_request:46950", "right": "pull_request:47057", "accept": false, - "reason": "Different model and failure mode: Mamba2 OOM vs DeepSeekV3.2 load error." + "reason": "Mamba2 runtime/OOM fix vs DeepSeek V3.2 test offload changes; different models and failures." }, { "left": "pull_request:46927", "right": "pull_request:47054", "accept": false, - "reason": "Different models and bugs: GLM46V test failure vs MiniMax-M3-VL load error." + "reason": "GLM46V input-id dtype handling is unrelated to MiniMax-M3-VL load/offload test setup." }, { "left": "pull_request:46927", "right": "pull_request:46951", "accept": false, - "reason": "Unrelated fixes in different subsystems/models." + "reason": "GLM46V input casting and PEFT loading are different code paths and different problems." }, { "left": "pull_request:46951", "right": "pull_request:46971", "accept": false, - "reason": "PEFT integration change vs Mamba2 test/loading fix." + "reason": "Checkpoint loading fix vs Mamba2 integration-test dtype/offload workaround; not the same change." }, { "left": "pull_request:46949", "right": "pull_request:46971", "accept": false, - "reason": "Whisper/audioflamingo cache fix vs Mamba2 OOM fix; not the same code path." + "reason": "Whisper cache-state initialization is unrelated to Mamba2 test loading changes." }, { "left": "pull_request:46951", "right": "pull_request:47057", "accept": false, - "reason": "Different subsystems and failure modes." + "reason": "PEFT loading behavior vs DeepSeek V3.2 offload test fixes are distinct issues." }, { "left": "pull_request:46941", "right": "pull_request:46970", "accept": false, - "reason": "Both are output-mismatch style PRs, but for different models and different code paths." + "reason": "DAC output-mismatch model fixes are unrelated to MusicGen Melody test expectations." }, { "left": "pull_request:47056", "right": "pull_request:47058", "accept": false, - "reason": "Whisper vs Mamba2; unrelated model bugs." + "reason": "Whisper cache initialization and Mamba2 OOM test tweaks do not fix the same bug." }, { "left": "pull_request:47006", "right": "pull_request:47058", "accept": false, - "reason": "Different models and distinct fixes." + "reason": "Gemma3 generation-cache logic and Mamba2 OOM test adjustments are different problems." }, { "left": "pull_request:46927", "right": "pull_request:47055", "accept": false, - "reason": "Different model families and different test/fix scopes." + "reason": "GLM46V input casting is unrelated to EdgeTam processor/backbone kwarg cleanup." }, { "left": "pull_request:46908", "right": "pull_request:47006", "accept": false, - "reason": "DeepSeekV3.2 load-error fix vs Gemma3 output-mismatch fix." + "reason": "DeepSeek V3.2 offload testing and Gemma3 generation-cache behavior are unrelated." }, { "left": "pull_request:46908", "right": "pull_request:46928", "accept": false, - "reason": "Different models and failure classes." + "reason": "DeepSeek V3.2 load/offload tests and BLIP2 tokenizer allowlisting are different changes." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46926", + "accept": false, + "reason": "EdgeTam post-processing kwarg handling is unrelated to MiniMax-M3-VL load/offload testing." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46941", + "accept": false, + "reason": "GLM46V input-id casting and DAC output-mismatch fixes target different models and bugs." }, { "left": "pull_request:46927", "right": "pull_request:46986", "accept": false, - "reason": "GLM46V vs GLM-OCR; related naming, but not the same bug or change." + "reason": "GLM46V input casting and GLM OCR output mismatch are separate model-specific fixes." + }, + { + "left": "pull_request:46986", + "right": "pull_request:47058", + "accept": false, + "reason": "GLM OCR output-shape mismatch and Mamba2 OOM handling are unrelated." }, { "left": "pull_request:46926", "right": "pull_request:46928", "accept": false, - "reason": "Different model failures and unrelated fixes." + "reason": "MiniMax-M3-VL offload testing and BLIP2 tokenizer registration are unrelated changes." }, { "left": "pull_request:46928", "right": "pull_request:47057", "accept": false, - "reason": "BLIP-2 test/import issue vs DeepSeekV3.2 load error." + "reason": "BLIP2 tokenizer allowlisting does not address DeepSeek V3.2 loading/offload issues." }, { "left": "pull_request:46926", "right": "pull_request:46941", "accept": false, - "reason": "Different models and bugs." + "reason": "MiniMax-M3-VL test offload and DAC output-mismatch fixes are different code paths." }, { "left": "pull_request:46908", "right": "pull_request:47058", "accept": false, - "reason": "Different models and failures." + "reason": "DeepSeek V3.2 load/offload tests and Mamba2 OOM fixes are unrelated." }, { "left": "pull_request:46928", "right": "pull_request:47058", "accept": false, - "reason": "BLIP-2 vs Mamba2; unrelated." + "reason": "BLIP2 tokenizer config and Mamba2 OOM test fixes are unrelated." }, { "left": "pull_request:46928", "right": "pull_request:47054", "accept": false, - "reason": "Different models and fixes." + "reason": "BLIP2 tokenizer allowlisting is unrelated to MiniMax-M3-VL load/offload test changes." }, { "left": "pull_request:46941", "right": "pull_request:46986", "accept": false, - "reason": "Dac vs GLM-OCR; no shared concrete bug." + "reason": "DAC output-mismatch fixes and GLM OCR output-mismatch fixes are different models and bugs." }, { "left": "pull_request:46941", "right": "pull_request:47006", "accept": false, - "reason": "Different model-specific output mismatches." + "reason": "DAC tensor-output fixes are unrelated to Gemma3 generation config behavior." }, { "left": "pull_request:46954", "right": "pull_request:46961", "accept": false, - "reason": "Unrelated: Pytest-v9 fp32 module change vs workflow cleanup." + "reason": "Pytest v9 fp32-module handling and Serge normalization workflow are unrelated changes." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46968", + "accept": false, + "reason": "Pytest v9 behavior fix is unrelated to docs autolink cleanup." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46974", + "accept": false, + "reason": "Workflow normalization automation and MusicGen generate typo fix are different changes." }, { "left": "pull_request:46958", - "right": "pull_request:46961", + "right": "pull_request:46974", "accept": false, - "reason": "CohereASR loss fix vs workflow normalization PR; not the same change." + "reason": "CohereASR loss handling and MusicGen generate typo are unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46971", + "accept": false, + "reason": "Pytest v9 fp32-module fix is unrelated to Mamba2 test loading/OOM changes." + }, + { + "left": "pull_request:46971", + "right": "pull_request:46974", + "accept": false, + "reason": "Mamba2 OOM test setup and MusicGen generate typo are unrelated." }, { "left": "pull_request:46958", + "right": "pull_request:46968", + "accept": false, + "reason": "CohereASR training-loss fix and docs autolink cleanup are unrelated." + }, + { + "left": "pull_request:46968", "right": "pull_request:46974", "accept": false, - "reason": "Different models and failure modes." + "reason": "Docs autolink cleanup and MusicGen generate typo fix are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46968", + "accept": false, + "reason": "Serge workflow automation and docs autolink cleanup are unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47078", + "accept": false, + "reason": "Both are Japanese docs PRs, but one adds DistilBERT docs while the other tweaks wording; not the same change." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46957", + "accept": false, + "reason": "DINO conversion/model modernization and finegrained FP8 generation changes do not address the same bug." + }, + { + "left": "pull_request:47019", + "right": "pull_request:47025", + "accept": false, + "reason": "Masking-utils causal-mask regression fix and Diffusion Gemma test updates are different scopes." + }, + { + "left": "pull_request:47062", + "right": "pull_request:47077", + "accept": false, + "reason": "IMDb link fix in sequence-classification docs is unrelated to Japanese DistilBERT docs translation." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47035", + "accept": false, + "reason": "Neuron training experiment scaffolding and torch-compile backend defaulting are related only by hardware, not the same fix." + }, + { + "left": "pull_request:47019", + "right": "pull_request:47069", + "accept": false, + "reason": "Mask-return contract regression and num_labels returning None are unrelated API changes." }, { "left": "pull_request:46957", "right": "pull_request:46965", "accept": false, - "reason": "Both touch DeepSeek V4, but one is a broad RoPE/test sweep and the other is a cache/use_cache behavior fix; not the same concrete bug." + "reason": "Finegrained FP8 generation optimization and DeepSeek V4 use_cache support are unrelated." }, { "left": "pull_request:46266", - "right": "pull_request:46995", + "right": "pull_request:47062", "accept": false, - "reason": "Dino modernization vs MLU flash-linear-attention availability; unrelated." + "reason": "DINO modernization and IMDb docs-link fixes are unrelated." }, { - "left": "pull_request:46707", - "right": "pull_request:46957", + "left": "pull_request:46912", + "right": "pull_request:47025", "accept": false, - "reason": "Core distributed/FSDP migration vs generation/quantization tuning; different change sets." + "reason": "RoPE test improvements and Diffusion Gemma test-case changes do not fix the same concrete issue." }, { - "left": "pull_request:46968", + "left": "pull_request:46912", + "right": "pull_request:47078", + "accept": false, + "reason": "RoPE test changes and Japanese docs wording updates are unrelated." + }, + { + "left": "pull_request:46196", "right": "pull_request:46995", "accept": false, - "reason": "Docs autolink cleanup vs hardware availability helper change; unrelated." + "reason": "wav2vec2 SDPA dispatch and flash-linear-attention MLU support are different backend-specific changes." }, { - "left": "pull_request:46912", + "left": "pull_request:46836", + "right": "pull_request:46995", + "accept": false, + "reason": "Unlimited OCR model addition is unrelated to flash-linear-attention availability checks." + }, + { + "left": "pull_request:46196", "right": "pull_request:47025", "accept": false, - "reason": "RoPE tests/model sweep vs DiffusionGemma test fix; different issues." + "reason": "wav2vec2 SDPA backend support and Diffusion Gemma tests are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47034", + "accept": false, + "reason": "DINO modernization and Neuron SFT experiment scripts are unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:47025", + "accept": false, + "reason": "Native FSDP2 migration and Diffusion Gemma test fixes are unrelated." + }, + { + "left": "pull_request:45694", + "right": "pull_request:47069", + "accept": false, + "reason": "Trainer batch-size split_batches fix and num_labels None typing change are different subsystems." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47079", + "accept": false, + "reason": "Molmo2 model addition and AltCLIP text embedding resize test are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46965", + "accept": false, + "reason": "Unlimited OCR model addition and DeepSeek V4 cache support are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46961", + "accept": false, + "reason": "Unlimited OCR model addition and Serge normalization workflow are unrelated." + }, + { + "left": "pull_request:46995", + "right": "pull_request:47003", + "accept": false, + "reason": "Flash-linear-attention MLU support and OpenVINO exporter work are different backend integrations." }, { "left": "pull_request:47035", "right": "pull_request:47069", "accept": false, - "reason": "Torch-compile backend default vs num_labels type/API fix; unrelated." + "reason": "Neuron compile-backend defaulting and num_labels return-type typing are unrelated." }, { - "left": "pull_request:47077", - "right": "pull_request:47078", + "left": "pull_request:46957", + "right": "pull_request:46961", "accept": false, - "reason": "Both are Japanese docs updates, but they are different documentation changes, not the same underlying patch." + "reason": "Finegrained FP8 generation work and Serge branch normalization workflow are unrelated." } ] }, "evaluator_result": null, "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -22266,9 +22334,8 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-47", + "cluster_id": "cluster-40546-48", "nodes": [ - "pull_request:40546", "pull_request:43451", "pull_request:44517", "pull_request:45694", @@ -22310,10 +22377,11 @@ "pull_request:47078" ], "soft_pairs": [ + "pull_request:46961|pull_request:47035", "pull_request:43451|pull_request:47025", - "pull_request:43451|pull_request:46266", "pull_request:46707|pull_request:47077", "pull_request:46912|pull_request:47077", + "pull_request:43451|pull_request:46266", "pull_request:46266|pull_request:47003", "pull_request:46196|pull_request:46961", "pull_request:46658|pull_request:46965", @@ -22344,8 +22412,8 @@ "pull_request:46968|pull_request:47003", "pull_request:46895|pull_request:46958", "pull_request:46907|pull_request:47055", - "pull_request:46950|pull_request:46971", "pull_request:46949|pull_request:47056", + "pull_request:46950|pull_request:46971", "pull_request:46950|pull_request:47058", "pull_request:46971|pull_request:47058", "pull_request:46926|pull_request:47054", @@ -22380,445 +22448,426 @@ "pull_request:43451|pull_request:46836", "pull_request:46836|pull_request:47003", "pull_request:43451|pull_request:47077", - "pull_request:46707|pull_request:46836", - "pull_request:40546|pull_request:43451" + "pull_request:46707|pull_request:46836" ], - "prepared_review_unit_hash": "938eac6bbeed3032a83e72f0c975a02cc5284680c218db1548ed31f4062b3f78", + "prepared_review_unit_hash": "2e2973c1b08c68a1b395072efa63082304dd1e5651893f271218830aaf4081cb", "component_budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 156, - "serialized_chars": 173692, - "estimated_input_tokens": 43423, - "estimated_eval_tokens": 87102 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 }, "budget": { - "node_count": 40, - "item_count": 40, + "node_count": 39, + "item_count": 39, "soft_pair_count": 72, - "serialized_chars": 118702, - "estimated_input_tokens": 29676, - "estimated_eval_tokens": 59608 + "serialized_chars": 115600, + "estimated_input_tokens": 28900, + "estimated_eval_tokens": 58056 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many soft matches are false positives across unrelated model-addition, docs, and integration-test PRs. The true duplicate sets are narrow and local: CohereASR loss-shift fixes, EdgeTam kwargs cleanup, Whisper cache-init fixes, DeepSeek/Minimax test offload fixes, and a few Mamba2 test-only OOM follow-ups.", - "confidence": 0.9, + "summary": "Mostly unrelated PRs were grouped together; only a few pairs are true duplicates on the same file/code path. The strongest duplicate signals are Edgetam forward-kwarg filtering, Whisper cache initialization, CohereASR loss shifting, and two test-only load_error fixes.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "46895 is the best anchor in the cluster because it is a concrete code-path fix with an explicit issue target, and 46958 is the closest true duplicate in the set.", + "canonical_pr_reason": "Pick pull_request:47055 as the canonical PR: it is the clearest and most complete Edgetam kwarg-filtering fix, matches 46907 on the same model files and forward path, and directly addresses the same concrete bug.", "best_issue_reason": null, - "best_pr_reason": "46895 is the strongest representative of the only clean duplicate family in the cluster; the rest of the items are mostly unrelated soft-similarity matches.", + "best_pr_reason": "pull_request:47055 is the best overall representative among the true duplicates because it fixes a concrete runtime code path in Edgetam rather than a broad subsystem or doc/test-only change, and it subsumes the smaller 46907 patch.", "soft_edge_verdicts": [ { - "left": "pull_request:43451", - "right": "pull_request:47025", + "left": "pull_request:46961", + "right": "pull_request:47035", "accept": false, - "reason": "Different model/docs changes; same broad docs/model pattern only." + "reason": "Unrelated changes: Serge workflow normalization vs torch.compile backend defaulting to neuron. Different code paths and goals." }, { "left": "pull_request:43451", - "right": "pull_request:46266", + "right": "pull_request:47025", "accept": false, - "reason": "Unrelated model-addition PRs for different models." + "reason": "Molmo2 model/docs addition is unrelated to DiffusionGemma test-case adjustments. Different models and problems." }, { "left": "pull_request:46707", "right": "pull_request:47077", "accept": false, - "reason": "Native FSDP2 runtime work vs Japanese docs translation; unrelated." + "reason": "Native FSDP2 module/migration is unrelated to Japanese DistilBERT docs translation. Not the same change." }, { "left": "pull_request:46912", "right": "pull_request:47077", "accept": false, - "reason": "RoPE/model test changes vs docs translation; unrelated." + "reason": "RoPE test coverage changes and Japanese docs wording are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46266", + "accept": false, + "reason": "Molmo2 addition and DINO conversion-mapping modernization are different model families and code paths." }, { "left": "pull_request:46266", "right": "pull_request:47003", "accept": false, - "reason": "Different subsystems: model modernization vs exporter docs/runtime." + "reason": "DINO conversion/model changes do not match the OpenVINO exporter feature work." }, { "left": "pull_request:46196", "right": "pull_request:46961", "accept": false, - "reason": "Wav2Vec2 SDPA dispatch vs workflow cleanup; unrelated." + "reason": "wav2vec2 SDPA dispatch support is unrelated to Serge branch normalization workflow plumbing." }, { "left": "pull_request:46658", "right": "pull_request:46965", "accept": false, - "reason": "Step-3.7 Flash build/convert work vs DeepSeek V4 cache support; unrelated." + "reason": "Step 3.7 Flash model conversion/setup work is unrelated to DeepSeek V4 cache handling." }, { "left": "pull_request:45694", "right": "pull_request:47035", "accept": false, - "reason": "TrainingArguments formatting/arg fix vs Neuron compile backend default; unrelated." + "reason": "Batch-size/split-batches behavior in TrainingArguments is unrelated to Neuron compile backend defaults." }, { "left": "pull_request:43451", "right": "pull_request:47078", "accept": false, - "reason": "Different documentation locales and model docs; unrelated." + "reason": "Molmo2 model addition is unrelated to Japanese docs wording cleanup." }, { "left": "pull_request:46658", "right": "pull_request:46961", "accept": false, - "reason": "Step-3.7 Flash tooling vs workflow normalization; unrelated." + "reason": "Step 3.7 Flash work and Serge workflow normalization are unrelated." }, { "left": "pull_request:46961", "right": "pull_request:46965", "accept": false, - "reason": "Workflow cleanup vs DeepSeek V4 cache support; unrelated." + "reason": "Workflow normalization and DeepSeek V4 cache behavior are unrelated changes." }, { "left": "pull_request:47003", "right": "pull_request:47078", "accept": false, - "reason": "OpenVINO exporter docs vs Japanese docs wording; unrelated." + "reason": "OpenVINO exporter feature work is unrelated to Japanese documentation wording." }, { "left": "pull_request:43451", "right": "pull_request:44517", "accept": false, - "reason": "Different model additions (VibeVoice vs Qwen3-TTS); same docs/auto-mapping pattern only." + "reason": "Molmo2 and Qwen3-TTS are different model additions with different files and tasks." }, { "left": "pull_request:45694", "right": "pull_request:46707", "accept": false, - "reason": "TrainingArguments fix vs FSDP2 module migration; unrelated." + "reason": "TrainingArguments split-batches fix is unrelated to native FSDP2 module migration." }, { "left": "pull_request:46836", "right": "pull_request:47078", "accept": false, - "reason": "Unlimited OCR model addition vs Japanese docs edits; unrelated." + "reason": "Unlimited OCR model addition is unrelated to Japanese docs wording." }, { "left": "pull_request:44517", "right": "pull_request:46836", "accept": false, - "reason": "Different model additions; shared registration/docs scaffolding is not enough." + "reason": "Qwen3-TTS and Unlimited OCR are separate model additions, not the same underlying change." }, { "left": "pull_request:46961", "right": "pull_request:47069", "accept": false, - "reason": "Workflow normalization vs num_labels typing/doc fix; unrelated." + "reason": "Workflow normalization and num_labels typing behavior are unrelated." }, { "left": "pull_request:44517", "right": "pull_request:47003", "accept": false, - "reason": "Qwen3-TTS model add vs OpenVINO exporter; unrelated." + "reason": "Qwen3-TTS model addition is unrelated to OpenVINO exporter support." }, { "left": "pull_request:43451", "right": "pull_request:46957", "accept": false, - "reason": "VibeVoice model add vs finegrained FP8 generation/runtime changes; unrelated." + "reason": "Molmo2 addition is unrelated to finegrained FP8 / decode optimization changes." }, { "left": "pull_request:43451", "right": "pull_request:46961", "accept": false, - "reason": "Model addition vs workflow cleanup; unrelated." + "reason": "Molmo2 model addition is unrelated to Serge workflow normalization." }, { "left": "pull_request:46961", "right": "pull_request:47003", "accept": false, - "reason": "Workflow cleanup vs OpenVINO exporter docs/runtime; unrelated." + "reason": "Workflow normalization and OpenVINO exporter support are unrelated." }, { "left": "pull_request:46961", "right": "pull_request:47034", "accept": false, - "reason": "Workflow normalization vs Neuron fine-tuning experiment; unrelated." + "reason": "Workflow normalization is unrelated to Neuron SFT experiment scripts and training args changes." }, { "left": "pull_request:46912", "right": "pull_request:46961", "accept": false, - "reason": "RoPE/model test work vs workflow cleanup; unrelated." + "reason": "RoPE test changes and workflow normalization are unrelated." }, { "left": "pull_request:46658", "right": "pull_request:46895", "accept": false, - "reason": "Step-3.7 Flash tooling vs CohereASR loss fix; unrelated." + "reason": "Step 3.7 Flash work and CohereASR loss-shift fix are unrelated model changes." }, { "left": "pull_request:44517", "right": "pull_request:47077", "accept": false, - "reason": "Different model docs translations; not the same change." + "reason": "Qwen3-TTS model addition is unrelated to Japanese DistilBERT docs translation." }, { "left": "pull_request:43451", "right": "pull_request:47069", "accept": false, - "reason": "Model addition vs configuration typing fix; unrelated." + "reason": "Molmo2 addition is unrelated to the num_labels property typing fix." }, { "left": "pull_request:43451", "right": "pull_request:45694", "accept": false, - "reason": "Model addition vs training args fix; unrelated." + "reason": "Molmo2 addition is unrelated to TrainingArguments split-batches behavior." }, { "left": "pull_request:43451", "right": "pull_request:46895", "accept": false, - "reason": "Model addition/docs vs CohereASR training-loss fix; unrelated." + "reason": "Molmo2 addition is unrelated to CohereASR training-loss fix." }, { "left": "pull_request:45694", "right": "pull_request:47003", "accept": false, - "reason": "TrainingArguments fix vs exporter docs/runtime; unrelated." + "reason": "TrainingArguments split-batches fix is unrelated to OpenVINO exporter support." }, { "left": "pull_request:46895", "right": "pull_request:47003", "accept": false, - "reason": "CohereASR loss fix vs exporter docs/runtime; unrelated." + "reason": "CohereASR loss-shift fix is unrelated to OpenVINO exporter support." }, { "left": "pull_request:46836", - "right": "pull_request:46895", + "right": "pull_request:47003", "accept": false, - "reason": "Unlimited OCR model addition vs CohereASR loss fix; unrelated." + "reason": "Unlimited OCR addition and OpenVINO exporter work are different feature areas." }, { "left": "pull_request:46968", "right": "pull_request:47003", "accept": false, - "reason": "Docs autolinks vs OpenVINO exporter docs/runtime; unrelated." + "reason": "Docs autolink cleanup is unrelated to OpenVINO exporter implementation." }, { "left": "pull_request:46895", "right": "pull_request:46958", "accept": true, - "reason": "Same CohereASR training-loss double-shift bug in the same forward/loss path." + "reason": "Same CohereASR training-loss double-shift fix: both touch the same model forward path and the same files, with the same issue target 46894." }, { "left": "pull_request:46907", "right": "pull_request:47055", - "accept": false, - "reason": "EdgeTam fix vs a different EdgeTam-related test/import triage item; not the same patch." - }, - { - "left": "pull_request:46950", - "right": "pull_request:46971", - "accept": false, - "reason": "Model-kernel OOM fix vs test-only dtype/offload tweak; different fixes." + "accept": true, + "reason": "Same Edgetam bug: both patches remove processor-only kwargs before calling the backbone in the same model files; 47055 is just the broader follow-up of 46907." }, { "left": "pull_request:46949", "right": "pull_request:47056", "accept": true, - "reason": "Same Whisper cache-init bug in the same attention forward path." + "reason": "Same Whisper cache-state bug: both change the same forward path in modeling_whisper.py to initialize cache update state consistently." }, { "left": "pull_request:46950", - "right": "pull_request:47058", + "right": "pull_request:46971", "accept": false, - "reason": "Model runtime OOM fix vs test harness OOM workaround; not the same underlying change." + "reason": "Both mention Mamba2 OOM, but one changes model execution and the other only adjusts test loading arguments; not the same concrete fix." }, { - "left": "pull_request:46971", + "left": "pull_request:46950", "right": "pull_request:47058", - "accept": true, - "reason": "Both are Mamba2 integration-test OOM mitigations in the same test file and can plausibly be merged together." + "accept": false, + "reason": "One PR changes Mamba2 kernel behavior, the other only tweaks integration tests and cleanup; different code paths." }, { "left": "pull_request:46926", "right": "pull_request:47054", "accept": true, - "reason": "Same MiniMax-M3-VL load_error/offload-folder fix in the same test loader path." + "reason": "Same MiniMax-M3-VL load_error workaround: both modify the same integration test to add an offload folder for device_map='auto' loading." }, { "left": "pull_request:46908", "right": "pull_request:47057", "accept": true, - "reason": "Same DeepSeek V3.2 load_error/offload-folder fix in the same integration test path." + "reason": "Same DeepSeekV3.2 load_error workaround: both patch the same test file to use a temporary offload folder for auto device-mapped loading." }, { "left": "pull_request:46949", "right": "pull_request:46950", "accept": false, - "reason": "Whisper/audioflamingo3 cache fix vs Mamba2 OOM fix; unrelated." + "reason": "Different models and different fixes: Whisper cache handling versus Mamba2 OOM behavior." }, { - "left": "pull_request:47054", - "right": "pull_request:47057", - "accept": false, - "reason": "Different model test failures (MiniMax-M3-VL vs DeepSeek V3.2); unrelated." - }, - { - "left": "pull_request:46949", - "right": "pull_request:46951", - "accept": false, - "reason": "Whisper/audioflamingo3 cache fix vs PEFT load mismatch fix; unrelated." - }, - { - "left": "pull_request:46907", - "right": "pull_request:46927", + "left": "pull_request:46927", + "right": "pull_request:46949", "accept": false, - "reason": "EdgeTam import_or_config fix vs GLM46V tensor dtype fix; unrelated." + "reason": "GLM46V input-id casting and Whisper cache initialization are unrelated model fixes." }, { "left": "pull_request:46908", "right": "pull_request:47054", "accept": false, - "reason": "DeepSeek V3.2 load_error fix vs MiniMax-M3-VL load_error fix; same symptom, different model/path." - }, - { - "left": "pull_request:46927", - "right": "pull_request:46971", - "accept": false, - "reason": "GLM46V dtype fix vs Mamba2 OOM workaround; unrelated." + "reason": "DeepSeekV3.2 and MiniMax-M3-VL are different model test fixes, even though both use tempfile/offload_folder." }, { "left": "pull_request:46926", "right": "pull_request:47057", "accept": false, - "reason": "MiniMax-M3-VL load_error fix vs DeepSeek V3.2 load_error fix; different model paths." + "reason": "MiniMax-M3-VL and DeepSeekV3.2 are different models and separate test workarounds." }, { - "left": "pull_request:46908", - "right": "pull_request:46926", + "left": "pull_request:46970", + "right": "pull_request:47006", "accept": false, - "reason": "DeepSeek V3.2 vs MiniMax-M3-VL test failures; unrelated." + "reason": "musicgen_melody test-output mismatch fix and gemma3 generation-config fix are unrelated." }, { "left": "pull_request:46986", "right": "pull_request:47006", "accept": false, - "reason": "Different model families and different generation/model fixes; not the same change." + "reason": "glm_ocr output mismatch tests and gemma3 generation behavior are unrelated." }, { "left": "pull_request:47054", "right": "pull_request:47055", "accept": false, - "reason": "MiniMax-M3-VL offload fix vs EdgeTam post-processing kwarg cleanup; unrelated." + "reason": "MiniMax-M3-VL test load workaround and Edgetam forward-path fix are different models and problems." }, { "left": "pull_request:46970", - "right": "pull_request:47006", + "right": "pull_request:47056", "accept": false, - "reason": "MusicGen-Melody output-mismatch fixes vs Gemma3 generation/cache fixes; unrelated." + "reason": "musicgen_melody and Whisper are separate model-specific fixes." }, { - "left": "pull_request:46970", - "right": "pull_request:46986", + "left": "pull_request:46971", + "right": "pull_request:47058", + "accept": false, + "reason": "Both are Mamba2-related, but one is a test loading tweak and the other is a model/kernel change; not the same underlying bug fix." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47003", "accept": false, - "reason": "Different model-specific output-mismatch fixes; not the same bug." + "reason": "Finegrained FP8 decode optimization work is unrelated to OpenVINO exporter support." }, { - "left": "pull_request:47057", - "right": "pull_request:47058", + "left": "pull_request:47055", + "right": "pull_request:47057", "accept": false, - "reason": "DeepSeek V3.2 load_error fix vs Mamba2 OOM workaround; unrelated." + "reason": "Edgetam forward-path kwarg filtering and DeepSeekV3.2 test loading work are unrelated." }, { "left": "pull_request:46907", "right": "pull_request:47054", "accept": false, - "reason": "EdgeTam fix vs MiniMax-M3-VL load_error fix; unrelated." + "reason": "Edgetam fix and MiniMax-M3-VL test workaround are different models and different code paths." }, { "left": "pull_request:46907", "right": "pull_request:47056", "accept": false, - "reason": "EdgeTam fix vs Whisper cache fix; unrelated." + "reason": "Edgetam and Whisper fixes target different model implementations." }, { - "left": "pull_request:46908", + "left": "pull_request:47054", "right": "pull_request:47056", "accept": false, - "reason": "DeepSeek V3.2 test load_error fix vs Whisper cache fix; unrelated." + "reason": "MiniMax-M3-VL and Whisper are unrelated model-specific fixes." }, { "left": "pull_request:47054", "right": "pull_request:47058", "accept": false, - "reason": "MiniMax-M3-VL offload fix vs Mamba2 OOM workaround; unrelated." + "reason": "MiniMax-M3-VL test workaround and Mamba2 OOM fix are different code paths." }, { "left": "pull_request:47055", "right": "pull_request:47057", "accept": false, - "reason": "EdgeTam kwarg cleanup vs DeepSeek V3.2 load_error fix; unrelated." + "reason": "Edgetam forward fix and DeepSeekV3.2 test workaround are unrelated." }, { "left": "pull_request:46907", "right": "pull_request:47006", "accept": false, - "reason": "EdgeTam fix vs Gemma3 generation/cache fix; unrelated." + "reason": "Edgetam kwarg filtering and gemma3 generation-config changes are unrelated." }, { "left": "pull_request:46907", "right": "pull_request:46970", "accept": false, - "reason": "EdgeTam fix vs MusicGen-Melody output-mismatch fix; unrelated." + "reason": "Edgetam fix and musicgen_melody output mismatch fix are unrelated model changes." }, { "left": "pull_request:46970", "right": "pull_request:47055", "accept": false, - "reason": "MusicGen-Melody output-mismatch fixes vs EdgeTam kwarg cleanup; unrelated." + "reason": "musicgen_melody test-output fix and Edgetam forward-path fix are unrelated." }, { "left": "pull_request:46907", "right": "pull_request:46986", "accept": false, - "reason": "EdgeTam fix vs GLM_OCR output-mismatch fix; unrelated." + "reason": "Edgetam forward-path cleanup and glm_ocr output mismatch tests are unrelated." }, { "left": "pull_request:46707", "right": "pull_request:47034", "accept": false, - "reason": "FSDP2 migration vs Neuron fine-tuning experiment; unrelated." + "reason": "Native FSDP2 migration and Neuron SFT experiment scaffolding are different initiatives." }, { "left": "pull_request:46836", "right": "pull_request:47062", "accept": false, - "reason": "Unlimited OCR model addition vs IMDb doc-link translation fix; unrelated." + "reason": "Unlimited OCR model addition and IMDb link documentation fix are unrelated." }, { "left": "pull_request:46658", "right": "pull_request:46836", "accept": false, - "reason": "Step-3.7 Flash tooling vs Unlimited OCR model addition; unrelated." + "reason": "Step 3.7 Flash model work and Unlimited OCR addition are unrelated model integrations." }, { "left": "pull_request:43451", "right": "pull_request:47077", "accept": false, - "reason": "VibeVoice model add vs Japanese DistilBERT docs translation; unrelated." + "reason": "Molmo2 model addition is unrelated to Japanese DistilBERT docs translation." }, { "left": "pull_request:46707", "right": "pull_request:46836", "accept": false, - "reason": "FSDP2 migration vs Unlimited OCR model addition; unrelated." - }, - { - "left": "pull_request:40546", - "right": "pull_request:43451", - "accept": false, - "reason": "Implement VibeVoice vs Molmo2 model addition; unrelated." + "reason": "FSDP2 migration and Unlimited OCR model addition are unrelated changes." } ] }, @@ -22840,8 +22889,9 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-47", + "cluster_id": "cluster-40546-48", "nodes": [ + "pull_request:40546", "pull_request:43451", "pull_request:46196", "pull_request:46658", @@ -22852,6 +22902,7 @@ "pull_request:47035" ], "soft_pairs": [ + "pull_request:40546|pull_request:43451", "pull_request:43451|pull_request:46196", "pull_request:46196|pull_request:47003", "pull_request:43451|pull_request:46658", @@ -22865,100 +22916,112 @@ "pull_request:46912|pull_request:47003", "pull_request:43451|pull_request:46707" ], - "prepared_review_unit_hash": "2fda3fd6cfaf7919b96fecd94458feb0657f325af36af1541384f6de30b71a8a", + "prepared_review_unit_hash": "cf508ebbcbf6f0cfd27856fe7da89a0e2d47ba730eec7c482ff01be4fbbfe97c", "component_budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 156, - "serialized_chars": 173692, - "estimated_input_tokens": 43423, - "estimated_eval_tokens": 87102 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 }, "budget": { - "node_count": 8, - "item_count": 8, - "soft_pair_count": 12, - "serialized_chars": 24884, - "estimated_input_tokens": 6221, - "estimated_eval_tokens": 12698 + "node_count": 9, + "item_count": 9, + "soft_pair_count": 13, + "serialized_chars": 28300, + "estimated_input_tokens": 7075, + "estimated_eval_tokens": 14406 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes model additions, exporter work, distributed/FSDP infrastructure, RoPE test expansion, and Neuron/training-args changes. None of the soft-edge pairs look like the same underlying bug or mergeable duplicate PR.", - "confidence": 0.98, + "summary": "This cluster is heterogeneous: it mixes unrelated model additions, exporter/runtime changes, training tweaks, and test-only work. I do not see any true duplicate PR pairs among them.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR stands out because the items are different features/fixes in unrelated code paths, not duplicate implementations of the same change.", + "canonical_pr_reason": "No clear canonical duplicate PR exists; the cluster does not center on one underlying change. If one had to represent the group, PR 43451 (Add Molmo2) is the most complete self-contained change, but it is not a duplicate target for the others.", "best_issue_reason": null, - "best_pr_reason": "No best PR for deduplication purposes; the cluster does not represent one underlying change.", + "best_pr_reason": "PR 43451 is the most complete and coherent single change in the set (model code, processors, tests, and docs). It is still not globally suitable as a duplicate representative because the rest of the cluster concerns different subsystems and different problems.", "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "Both are model-addition PRs with docs/auto-mapping churn, but they implement different models (VibeVoice vs Molmo2) and different code paths." + }, { "left": "pull_request:43451", "right": "pull_request:46196", "accept": false, - "reason": "Molmo2 model onboarding vs wav2vec2 SDPA dispatch support are unrelated changes in different model families." + "reason": "Molmo2 model integration and wav2vec2 SDPA dispatch are unrelated changes; overlap in audio/model infrastructure is too broad to be a duplicate." }, { "left": "pull_request:46196", "right": "pull_request:47003", "accept": false, - "reason": "wav2vec2 attention dispatch and OpenVINO exporter support are different subsystems and different code paths." + "reason": "One adds SDPA dispatch for audio models; the other adds an OpenVINO exporter. Different goals, subsystems, and merge scopes." }, { "left": "pull_request:43451", "right": "pull_request:46658", "accept": false, - "reason": "Molmo2 addition and Step-3.7-Flash model/conversion work are separate model integrations, not the same fix." + "reason": "Molmo2 integration and Step-3.7 Flash integration are different model families and different implementations, not one underlying bug/change." }, { "left": "pull_request:46658", "right": "pull_request:47003", "accept": false, - "reason": "Step-3.7-Flash infrastructure and OpenVINO exporter work address unrelated functionality." + "reason": "Step-3.7 Flash implementation and OpenVINO exporter work are unrelated; they do not fix the same concrete code-path problem." }, { "left": "pull_request:43451", "right": "pull_request:47003", "accept": false, - "reason": "A new model integration is not a duplicate of exporter support for OpenVINO." + "reason": "Molmo2 model addition vs exporter support are distinct changes; shared docs/auto files are incidental." }, { "left": "pull_request:43451", "right": "pull_request:47034", "accept": false, - "reason": "Molmo2 model addition and Neuron SFT scripts/training tweaks are unrelated." + "reason": "Molmo2 addition and the Neuron SFT experiment script address different tasks and code paths." }, { "left": "pull_request:43451", "right": "pull_request:47035", "accept": false, - "reason": "Molmo2 model onboarding and a torch-compile backend default for Neuron are different changes." + "reason": "Molmo2 model addition is unrelated to the torch-compile backend default change for Neuron." }, { "left": "pull_request:43451", "right": "pull_request:46912", "accept": false, - "reason": "A model addition and RoPE test/implementation changes do not share the same underlying bug or feature." + "reason": "Molmo2 integration and RoPE test expansion are different kinds of changes; the overlap is only at a very broad framework level." }, { "left": "pull_request:47003", "right": "pull_request:47034", "accept": false, - "reason": "OpenVINO exporter support and Neuron finetuning scripts target different export/training workflows." + "reason": "OpenVINO exporter support and Neuron fine-tuning scripts are different runtime features with no shared concrete bug." }, { "left": "pull_request:47003", "right": "pull_request:47035", "accept": false, - "reason": "Exporter work and torch-compile backend defaulting for Neuron are unrelated code paths." + "reason": "Exporter support and Neuron compile-backend defaulting are separate infrastructure changes, not the same fix." }, { "left": "pull_request:46912", "right": "pull_request:47003", "accept": false, - "reason": "RoPE test improvements are not the same change as adding an OpenVINO exporter." + "reason": "RoPE testing improvements do not address the exporter change; these are unrelated maintenance areas." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46707", + "accept": false, + "reason": "Molmo2 model addition and native FSDP2 migration are different large-scale changes with no common underlying issue." } ] }, @@ -22988,7 +23051,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "6aeb0e269fe4276653b710d6dda770add1f72a4fc04a11681df1193e7cfa885d", + "prepared_review_unit_hash": "338299dbf3edd54e6e21fdf23242105d61bbac0079210ac643368602585e733a", "component_budget": { "node_count": 2, "item_count": 2, @@ -23010,18 +23073,18 @@ "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.", + "summary": "False positive cluster: one PR adds TimesFM covariate forecasting, the other adds DINOv3 image classification support. They touch different models, code paths, and user-facing features.", "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the PRs modify different code paths, model families, docs, and tests, so there is no single underlying change to deduplicate.", + "canonical_pr_reason": "No canonical PR: these are unrelated feature additions in different subsystems, not the same underlying change.", "best_issue_reason": null, - "best_pr_reason": "Neither PR is a good global representative for the cluster because they address distinct features rather than the same bug or change.", + "best_pr_reason": "No best PR for the cluster; neither PR is a suitable representative of the other because they do not overlap in purpose or implementation.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different underlying changes: 40755 adds TimesFM forecasting with covariates across TimesFM code/docs/tests, while 41224 adds DINOv3 image classification support. They do not fix the same code-path problem and would not be merged into one PR." + "reason": "TimesFM covariate forecasting and DINOv3 image classification are unrelated changes with no shared bug or code-path fix." } ] }, @@ -23142,12 +23205,12 @@ "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "963510786e397a0cbea16fbabddec30af985eb0c73417e8646a6b540f37106a8", + "prepared_review_unit_hash": "936ef4bf90712da975974bde757cbcc8592d8d4488e12bcb6fbec39bb3ccd123", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 34782, + "serialized_chars": 34781, "estimated_input_tokens": 8696, "estimated_eval_tokens": 17648 }, @@ -23155,7 +23218,7 @@ "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 34782, + "serialized_chars": 34781, "estimated_input_tokens": 8696, "estimated_eval_tokens": 17648 }, @@ -23164,126 +23227,126 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only one pair looks like a real duplicate: the two `fixed_cross_entropy` PRs against `loss_utils.py`. The rest are merely theme-adjacent (same broad subsystem or different model/feature work) and should not be merged as duplicates.", - "confidence": 0.94, + "summary": "This cluster is mostly unrelated PRs that only share broad soft-similarity signals. The only clear duplicate-like pair is the two fixed_cross_entropy PRs (43251 and 43254), which both change the same function in the same file for the same issue.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is the better canonical representative of the duplicate pair: it directly targets issue 43240 and makes the concrete `cross_entropy` forwarding fix in `loss_utils.py`.", + "canonical_pr_reason": "No single canonical PR for the whole cluster; most items are unrelated. If a representative is needed, PR 43251 is the stronger canonical for the only true duplicate pair because it explicitly targets issue 43240 and includes the fuller cross-entropy kwargs support.", "best_issue_reason": null, - "best_pr_reason": "PR 43251 is the best overall representative because it is the clearest, issue-linked fix for the actual bug in `fixed_cross_entropy`, while the competing PR is just an alternate formulation of the same change.", + "best_pr_reason": "PR 43251 is the best representative of the cluster\u2019s only genuine duplicate pair: it fixes the same cross-entropy behavior as 43254, but is more complete and explicitly tied to the tracking issue.", "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." + "reason": "Different features: Mistral tokenizer missing methods vs Executorch dynamic-shape export tests. No shared bug or change." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Unrelated changes to tokenizer methods vs adding MobileLLM model support." + "reason": "Tokenizer helper additions are unrelated to adding MobileLLM model files." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness and a new model addition are different changes." + "reason": "Processor auto-loading robustness changes are unrelated to adding MobileLLM." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Cross-entropy helper kwargs vs MobileLLM model implementation; no shared bug/change." + "reason": "Cross-entropy kwargs passthrough and a new model implementation are unrelated changes." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Loss helper forwarding fix and Mistral4 test/model work are unrelated." + "reason": "Loss utility kwargs support does not match Mistral4 test fixes; different code paths and purposes." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "ExecuTorch dynamic-shape export tests are unrelated to Mistral4 test fixes." + "reason": "Executorch export tests and Mistral4 fixes are unrelated." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Different code paths and features: export testing vs new model support." + "reason": "Dynamic-shape export testing is unrelated to MobileLLM model addition." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "Adding MobileLLM and fixing `cls.config` resolution are unrelated changes." + "reason": "New MobileLLM model code and config-resolution ordering fix are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "New model support vs Falcon output-interface refactor; not the same underlying change." + "reason": "MobileLLM model addition and Falcon output-collection refactor are unrelated." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor and model-creation skill scaffolding are different scopes." + "reason": "Falcon refactor and model-creation skill/docs are unrelated." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon forward-interface refactor and Mistral4 test fixes do not address the same bug." + "reason": "Falcon refactoring and Mistral4 test fixes address different models and problems." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test stabilization and CLIP-like `layrnorm` renaming are unrelated." + "reason": "Mistral4 tests and layrnorm->layernorm rename/conversion changes are different issues." }, { "left": "pull_request:44827", "right": "pull_request:46267", "accept": false, - "reason": "Mistral4 tests and chat response parsing updates are unrelated subsystems." + "reason": "Mistral4 test fixes and chat parsing updates are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 testing work and adding a TDT loss kernel are different features." + "reason": "Mistral4 test fixes and TDT loss kernel integration are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test fixes and config-resolution ordering fixes are unrelated changes." + "reason": "Mistral4 test fixes and config resolution ordering are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 test fixes and the model-creation skill PR are not the same underlying change." + "reason": "The shared auto/modeling_auto file is incidental; model registration for Sarvam MoE and Mistral4 tests are unrelated changes." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "A new model implementation and a model-creation skill are distinct artifacts, not duplicate fixes." + "reason": "Adding MobileLLM and creating a model-addition skill are unrelated." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs patch the same `src/transformers/loss/loss_utils.py` code path to forward supported `cross_entropy` kwargs; they are alternate fixes for the same bug and could plausibly be merged into one PR." + "reason": "Both PRs modify the same fixed_cross_entropy function to pass supported kwargs through to nn.functional.cross_entropy for the same underlying issue 43240; they could plausibly be merged into one patch." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary stays grounded in the packet: the only clearly duplicate pair is the two `fixed_cross_entropy` PRs in `src/transformers/loss/loss_utils.py`, and the rest are conservatively treated as theme-adjacent rather than duplicates. The accepted pair is plausibly mergeable into a single bugfix PR, and none of the rejected pairs are overstated as the same change." + "feedback": "The summary is well grounded in the packet, and the conservative non-match verdicts are supported by the titles/diffs. The 43251/43254 pair is the only clearly mergeable duplicate-like pair: both touch the same fixed_cross_entropy function for the same issue and the patches are compatible as a single bugfix." }, "evaluator_used": true, "retried": false, @@ -23310,7 +23373,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "8ceabc7cfe7b1f0ce49828cc5f94d07bdecabede61563396f5e895ba91ba64f0", + "prepared_review_unit_hash": "d1ded7571cd9eeb46327d8a6544cc3ff9c3602bcc35d1ef9d4a11cd474a71259", "component_budget": { "node_count": 2, "item_count": 2, @@ -23332,18 +23395,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs both update GitHub Actions workflow dependencies, but they address different components: one is a broad Node 24 compatibility bump for core actions, the other is a Docker build-action version upgrade. They overlap in workflow files, not in the underlying change.", - "confidence": 0.97, + "summary": "Two open PRs both update GitHub Actions dependencies, but they target different action families and different maintenance goals: Node 24 compatibility vs general/latest-version upgrades. They are not the same underlying change.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the broader workflow-actions compatibility update and is more representative of the Node 24 migration work; PR 42976 is a separate Docker build-action upgrade.", + "canonical_pr_reason": "PR 42975 is the cleaner representative because it specifically addresses the concrete Node 24 compatibility migration across workflow actions.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the better representative PR for general GitHub Actions maintenance, but it should not be treated as a duplicate of PR 42976 because the concrete changes differ.", + "best_pr_reason": "PR 42975 is the best representative of this cluster since its scope is more clearly defined and tied to a single compatibility objective; PR 42976 is a separate broader dependency bump.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Reject: these are not the same fix. 42975 upgrades checkout/cache/upload-artifact for Node 24 compatibility across workflows, while 42976 bumps docker/build-push-action for Docker image workflows. Shared workflow filenames alone are not enough to merge them as duplicates." + "reason": "Both are workflow dependency upgrades, but they change different actions and fix different problems. 42975 updates checkout/cache/upload-artifact for Node 24 compatibility, while 42976 updates docker/build-push-action and other actions to newer versions. They would not plausibly merge into one single PR as the same bug/change." } ] }, @@ -23378,12 +23441,12 @@ "pull_request:43448|pull_request:45490", "pull_request:45490|pull_request:46562" ], - "prepared_review_unit_hash": "7c6b6278cd2ba765ffc9929369e596aded56a5d5db5b5801423060e135a05155", + "prepared_review_unit_hash": "c45d3406d4ad45a8cfd39329f01f87f7decda1a50bc9e64a825ca719e2ec9bb3", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13046, + "serialized_chars": 13045, "estimated_input_tokens": 3262, "estimated_eval_tokens": 6780 }, @@ -23391,7 +23454,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13046, + "serialized_chars": 13045, "estimated_input_tokens": 3262, "estimated_eval_tokens": 6780 }, @@ -23400,36 +23463,36 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four pull requests are separate model-addition efforts; they share only generic Transformers boilerplate and auto-registry/docs updates, not the same underlying bug or change.", - "confidence": 0.97, + "summary": "All four pull requests are separate model-addition efforts for different architectures (ViT NEPA, Molmo, CTSM, Step3p7) and only overlap on generic docs/auto-registry plumbing. They are not duplicates of the same underlying change.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single PR is a clear canonical duplicate target; the cluster contains distinct model integrations rather than one repeated fix/change.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "None of these is a better duplicate representative than the others, because each introduces a different model and code path.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different models and codepaths: ViT NEPA vs Molmo. Shared docs/auto-mapping boilerplate is not enough to treat them as one PR." + "reason": "Different model additions (ViT NEPA vs Molmo). Shared registry/docs files are generic scaffolding, not evidence of the same code change." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Unrelated additions: ViT NEPA versus CTSM. The overlap is only framework plumbing, not the same concrete change." + "reason": "ViT NEPA and CTSM are unrelated model implementations; overlap is limited to common auto/docs wiring." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are distinct model integrations with different implementations and tests; not a mergeable duplicate." + "reason": "Molmo and CTSM add different models with distinct configs, processors, and tests; not the same underlying bug or change." }, { "left": "pull_request:45490", "right": "pull_request:46562", "accept": false, - "reason": "CTSM and Step3p7 are different model feature PRs. Similar auto/doc updates do not indicate the same underlying work." + "reason": "CTSM and Step3p7 are separate model integrations; similarity is only at the framework/registry level, so they should not be merged as duplicates." } ] }, @@ -23478,130 +23541,130 @@ "pull_request:46539|pull_request:46922", "pull_request:46886|pull_request:46922" ], - "prepared_review_unit_hash": "737c50bdc5f500ba5a85865d04b1778ecab5a91d5ff02848c0c034acb5e83ead", + "prepared_review_unit_hash": "137c93bc9e9558d4ceb8f73e9db30e05efb5f5bb6abdb92476ce95a24be5648e", "component_budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19599, - "estimated_input_tokens": 4900, - "estimated_eval_tokens": 10056 + "serialized_chars": 19595, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 }, "budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19599, - "estimated_input_tokens": 4900, - "estimated_eval_tokens": 10056 + "serialized_chars": 19595, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Most items are unrelated PRs with generic titles; the only real duplicate-style pair is the two YOLOS fixes for the same device_map mismatch path.", - "confidence": 0.98, + "summary": "This is a heterogeneous set of PRs: most are unrelated test/CI/model fixes across different subsystems, and only the two YOLOS PRs appear to address the same concrete device-map mismatch bug.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:46886 is the clearest representative: it has the explicit issue target, a specific title, and a focused fix for the YOLOS device mismatch bug.", + "canonical_pr_reason": "No single canonical PR for the whole set; the only clear duplicate pair is the YOLOS fix, where 46886 is the better anchor because it explicitly targets issue 46869 and describes the device_map=\"auto\" mismatch clearly.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46886 is the best cluster representative because it is the most precise and directly tied to the underlying YOLOS bug, while the others are either unrelated or broader/test-only.", + "best_pr_reason": "PR 46886 is the strongest representative candidate: focused YOLOS bug fix, explicit issue link, and the same underlying problem as 46922.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Both are vague test-related PRs with different models and different changes; the similarity is only the generic \"fix\" style." + "reason": "Different models and different changes; one is Whisper tokenization/tests, the other is CLIP/ViT test edits." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper tokenizer test fix vs FP8 dequant/loading fix; completely different code paths and bugs." + "reason": "Whisper decode behavior vs FP8 dequant/loading fixes; unrelated code paths." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "Whisper test cleanup and generation/watermarking/cis fix are unrelated changes in different subsystems." + "reason": "Both mention Whisper, but the concrete fixes differ: tokenizer timestamp/batch decode vs generation watermarking validation and broader CI/test updates." }, { "left": "pull_request:43291", "right": "pull_request:46922", "accept": false, - "reason": "Whisper tokenizer/test issue and YOLOS device mismatch are unrelated model-specific fixes." + "reason": "Whisper tokenizer fix and YOLOS device-mismatch fix are unrelated." }, { "left": "pull_request:43665", "right": "pull_request:46539", "accept": false, - "reason": "Model test failures in CLIP/ViT are not the same underlying issue as the Phimoe CI/runtime memory-limit fix." + "reason": "Separate test/CI failures in different model suites; no shared underlying bug." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Different models and different failure modes; one is test-only, the other is FP8 loading/dequant behavior." + "reason": "CLIP/ViT test edits are unrelated to FP8 dequant/loading changes." }, { "left": "pull_request:43665", "right": "pull_request:46380", "accept": false, - "reason": "CLIP/ViT test adjustments are unrelated to generation config validation and watermarking conversion." + "reason": "Different fixes in different areas; only broad test churn overlaps, not the same bug." }, { "left": "pull_request:43665", "right": "pull_request:46922", "accept": false, - "reason": "These touch different models and different bugs; same generic title style is not enough to merge them." + "reason": "Unrelated model test changes versus YOLOS runtime/device placement fix." }, { "left": "pull_request:44815", "right": "pull_request:46539", "accept": false, - "reason": "FP8 quantization/loading changes and Phimoe integration test CI fixes do not address the same code-path problem." + "reason": "FP8 dequant/loading work is unrelated to Phimoe integration test CI adjustments." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "Unrelated fixes: FP8 dequant logic versus generation/watermarking validation." + "reason": "Different subsystems and concrete bugs; shared test churn is not enough to merge them." }, { "left": "pull_request:44815", "right": "pull_request:46922", "accept": false, - "reason": "FP8 quantization/loading and YOLOS device placement are distinct bugs in different areas." + "reason": "FP8 dequantization changes and YOLOS device-map fixes are unrelated." }, { "left": "pull_request:46380", "right": "pull_request:46539", "accept": false, - "reason": "Generation config validation changes are unrelated to CI resource patching for Phimoe tests." + "reason": "Generation/watermarking and Whisper changes do not match the Phimoe CI-memory fix." }, { "left": "pull_request:46380", "right": "pull_request:46922", "accept": false, - "reason": "Different subsystems and different failures; not the same underlying change." + "reason": "Broad generation/Whisper validation changes are unrelated to YOLOS device placement." }, { "left": "pull_request:46539", "right": "pull_request:46922", "accept": false, - "reason": "Phimoe CI/device-map test infrastructure fix is unrelated to YOLOS runtime device mismatch." + "reason": "CI memory-limit/test harness changes do not address the YOLOS runtime bug." }, { "left": "pull_request:46886", "right": "pull_request:46922", "accept": true, - "reason": "Both target the same YOLOS device_map=\"auto\" device mismatch issue in the same file; 46922 adds an extra device move but could plausibly be merged into the same fix." + "reason": "Both PRs fix the same YOLOS device_map=\"auto\" device mismatch in the same file and could plausibly be merged into one change." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only mergeable same-bugfix pair is the two YOLOS PRs sharing issue target 46869 and the same file; the other pairs are reasonably treated as distinct changes in different subsystems." + "feedback": "The summary is grounded in the packet and stays conservative overall. The rejected pairs are appropriately treated as unrelated despite superficial filename/model overlap, and the only accepted pair (46886/46922) is supported by the shared YOLOS file and shared issue target, with changes that look mergeable into one bugfix PR. The canonical/representative wording is cautious enough." }, "evaluator_used": true, "retried": false, @@ -23631,52 +23694,52 @@ "pull_request:43488|pull_request:43532", "pull_request:43532|pull_request:46747" ], - "prepared_review_unit_hash": "e558b1f716a65659279b6a7e6871209204c88b0f71cd0b6f3fee0b3ad71f43c2", + "prepared_review_unit_hash": "7ba2a46a9e89bfa4ab730471688e5027d5ac7fe75419ec8a21b9cd2c71aed30a", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6348, - "estimated_input_tokens": 1587, - "estimated_eval_tokens": 3430 + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6348, - "estimated_input_tokens": 1587, - "estimated_eval_tokens": 3430 + "serialized_chars": 6349, + "estimated_input_tokens": 1588, + "estimated_eval_tokens": 3432 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three distinct \u201c[don\u2019t merge]\u201d PRs for different repo-bot/workflow/test tweaks, not duplicates of the same underlying change.", - "confidence": 0.89, + "summary": "All three pull requests are unrelated, intentionally marked \"don't merge,\" and appear to be test/bot/workflow demonstrations rather than the same underlying fix. None should be clustered as duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No PR is a true duplicate of the others, but #43532 is the broadest and most representative of the workflow/test-harness area, so it is the best single cluster representative if one must be chosen.", + "canonical_pr_reason": "No canonical PR: each PR targets a different file set and a different synthetic/demo change, so they do not represent one underlying fix or change.", "best_issue_reason": null, - "best_pr_reason": "#43532 has the richest context and the widest surface area across workflow automation, testing utilities, and a model test, making it the most useful representative of this non-duplicate cluster.", + "best_pr_reason": "No best PR: there is no shared bug or merged change to preserve, and all three look like separate test PRs rather than duplicate candidates.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:46747", "accept": false, - "reason": "Different concrete changes: #43488 is about packaging/versioning and a Vi test, while #46747 only changes the PR security-gate workflow trigger. Same broad repo area, but not the same fix." + "reason": "Different code paths and purposes: one changes setup/versioning and a Vit test file, the other changes a workflow trigger. They are not the same fix." }, { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "#43488 tweaks setup/dependency metadata and a test file; #43532 rewires several CI/workflow files plus testing utilities. They do not fix the same bug or change the same code path." + "reason": "Overlap is only superficial (shared test file mention). The PRs touch unrelated workflows/utilities versus packaging/version changes, so they cannot be one merged PR." }, { "left": "pull_request:43532", "right": "pull_request:46747", "accept": false, - "reason": "#43532 is a broad workflow/comment-ci refactor with test harness changes, whereas #46747 is a single-workflow trigger change for the security gate. Related subsystem, but not one mergeable duplicate change." + "reason": "One is a broad workflow/test-diff showcase, the other is a single security-gate workflow trigger change. No concrete shared bug or change." } ] }, @@ -23706,7 +23769,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "c09399cc8967c6f60d5fd663c3ce14540f898d2abc73c7702a90e98eff9714a7", + "prepared_review_unit_hash": "2cc3f5fda8ccd65260a5455877c423fbaa6ae6bb3cf4677ae22dd6c2c77a35df", "component_budget": { "node_count": 2, "item_count": 2, @@ -23728,18 +23791,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related only at the framework level: both add new segmentation pipelines and register them in shared docs/metadata, but they implement different features and different pipeline files/models. They should not be deduplicated.", + "summary": "These are two related but distinct PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They share some documentation and pipeline plumbing, but they target different tasks/models and should not be deduplicated.", "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR for duplication: the two PRs implement different segmentation features and do not appear to be the same underlying change.", "best_issue_reason": null, - "best_pr_reason": "PR 43613 is the broader representative of the visual-segmentation addition, but it is still a distinct feature from 43612 and not a duplicate candidate.", + "best_pr_reason": "PR #43613 is the broader, more central change in this cluster: it adds the visual segmentation pipeline with multiple model integrations and tests, making it the better representative PR if one must be chosen.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different underlying changes: 43612 adds Promptable Concept Segmentation, while 43613 adds Promptable Visual Segmentation. They share registration/docs plumbing, but the actual pipeline modules and model integrations are different, so they should not be merged as one PR." + "reason": "Reject: these are different pipeline additions (concept segmentation vs visual segmentation) with different target code paths and model/test coverage; they are related in theme but not the same underlying bug or change." } ] }, @@ -23769,7 +23832,7 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "b39a1ab4a904dd43a1c8e569705d43ab31b2610e82932a8de6800f76628f8f4e", + "prepared_review_unit_hash": "6cdac516889031a3b6ca966c9d437163ed5d78ad9ea201429a3dd9beb5c1e755", "component_budget": { "node_count": 2, "item_count": 2, @@ -23791,18 +23854,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs touch the same file but address different trainer concerns: 43651 changes gradient-accumulation loss scaling in `training_step`, while 43663 extracts signature-column logic for model input filtering. They are not the same underlying change.", - "confidence": 0.95, + "summary": "The two PRs both modify `src/transformers/trainer.py`, but they address different trainer extension points: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom trainers. They are not the same underlying change.", + "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 43651 is the more central trainer behavior change, but the two PRs are still distinct and should not be merged as duplicates.", + "best_pr_reason": "PR #43651 is the stronger standalone change in the cluster because it fixes a concrete training-step behavior and introduces a focused hook for custom trainers; however, it is still unrelated to #43663 and should not be treated as a duplicate.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Different code paths and purposes: loss scaling vs. signature-column filtering. Sharing `trainer.py` is too broad to treat as the same fix." + "reason": "Different code paths and different problems: #43651 is about gradient accumulation loss scaling in `training_step`, while #43663 is about extracting signature-column logic for dataset column filtering. Shared file alone is not enough to treat them as one PR." } ] }, @@ -23832,7 +23895,7 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "dd6bda6feff49ad8730af3c59c712435473013f4d7bd8e3c70abc9324161b9e2", + "prepared_review_unit_hash": "7474a49f668159de578d1030e6826bd8d809f0e71ad812f418f6012f4a389214", "component_budget": { "node_count": 2, "item_count": 2, @@ -23854,18 +23917,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same CLI file and mention TypeAdapter, but they appear to fix different NameError scenarios. PR 43842 is the more focused pydantic-absence fix; PR 43656 looks like a separate typing/annotation cleanup in the same module.", - "confidence": 0.87, + "summary": "Both PRs are in `src/transformers/cli/serve.py` and are about TypeAdapter/annotation-related import safety, but they fix different concrete failure modes and are tied to different issues.", + "confidence": 0.69, "canonical_issue_reason": null, - "canonical_pr_reason": "43842 is the better canonical PR because it explicitly targets the concrete crash mode in the title and diff: TypeAdapter NameError when pydantic is not installed, with a minimal, focused patch.", + "canonical_pr_reason": "PR 43656 is the broader, more representative fix for the CLI annotation NameError theme and has the stronger discussion/context around the file-wide typing cleanup.", "best_issue_reason": null, - "best_pr_reason": "43842 is the best representative of the cluster because it addresses a specific user-facing failure mode with the clearest and smallest change set.", + "best_pr_reason": "PR 43656 is the better standalone candidate because it addresses the import-time annotation evaluation problem across the file, not just one annotated parameter.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Same file and similar wording, but different fixes: 43656 changes annotation behavior/import typing cleanup, while 43842 specifically guards TypeAdapter behind TYPE_CHECKING for the missing-pydantic case. They are not the same underlying bug/change." + "reason": "Related, but not the same underlying bug/change: 43656 postpones annotation evaluation across `serve.py`, while 43842 specifically makes `TypeAdapter` typing safe when pydantic is absent. They target different issue reports and would not be a clean single merged PR." } ] }, @@ -24058,130 +24121,130 @@ "pull_request:44024|pull_request:44076", "pull_request:44073|pull_request:44074" ], - "prepared_review_unit_hash": "ececf8b82bbfc60e05b5ffc6b391315a0c2da4f36dda1e3de9468414d407539f", + "prepared_review_unit_hash": "7812f97bf46c537dbd306616e8375a11470b98a23fc1561868783d534434c0ed", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 76, - "serialized_chars": 84112, - "estimated_input_tokens": 21028, - "estimated_eval_tokens": 42312 + "serialized_chars": 84115, + "estimated_input_tokens": 21029, + "estimated_eval_tokens": 42314 }, "budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 72, - "serialized_chars": 76934, - "estimated_input_tokens": 19234, - "estimated_eval_tokens": 38724 + "serialized_chars": 76937, + "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 broad family of model-specific output-tracing refactors, not one duplicate PR. The only clear duplicate-like pairs are within the same GPT-J/CodeGen and GPT-Neo code paths; most cross-model pairs are separate changes and should be rejected.", - "confidence": 0.84, + "summary": "This cluster is mostly a theme of standardized output-tracing refactors across many unrelated model families. I only see clear duplicate PRs inside the GPT-J/codegen line, plus a GPT-Neo revision pair; the cross-model pairs should stay separate.", + "confidence": 0.87, "canonical_issue_reason": null, - "canonical_pr_reason": "44066 is the best cluster anchor: it is a focused, explicit output-tracing refactor with a tracking link and concrete GPT-J/CodeGen code-path changes, making it the clearest representative of the shared pattern.", + "canonical_pr_reason": "PR 44722 is the strongest representative of the duplicate GPT-J/codegen line: it is the latest and most complete variant, and it overlaps the same concrete gptj/modeling_gptj.py and codegen/modeling_codegen.py output-tracing changes as the earlier GPT-J PRs.", "best_issue_reason": null, - "best_pr_reason": "44066 is the strongest representative of the recurring standardized output-tracing work. It is explicit, concrete, and broad enough to summarize the core refactor without conflating unrelated model-specific PRs.", + "best_pr_reason": "PR 44722 is the best overall representative because it appears to supersede the earlier GPT-J/codegen output-tracing refactors rather than being a narrower or older partial version.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families and code paths (FNet/CVT vs GPT-J). Same high-level refactor pattern, but not the same underlying change." + "reason": "Different model families (FNet/CVT vs GPT-J); same general refactor pattern, but not the same underlying bug/change." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Different model families and implementations (FNet/CVT vs DeBERTaV2). Not a duplicate change." + "reason": "Unrelated architectures and code paths; both standardize outputs, but they are not duplicates." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Same GPT-J output-tracing refactor on the same code path; these are plausibly one combined PR." + "reason": "Same GPT-J output-tracing refactor on overlapping gptj/modeling_gptj.py changes; 44085 looks like a narrower/earlier variant of the same work." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "Both are output-tracing refactors, but they target different backbones (ResNet/RT-DETR ResNet vs EfficientNet). Not the same bug/change." + "reason": "Different models (RegNet vs EfficientNet); same refactor theme only, not the same change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet and GPT-J/CodeGen are unrelated code paths; same refactor theme is too broad to call them duplicates." + "reason": "EfficientNet and GPT-J/codegen are unrelated model implementations; no shared code-path bug." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen vs EfficientNet are unrelated implementations. Not the same concrete change." + "reason": "Different model families and distinct files; these are separate refactors." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTaV2 are different model families with different output paths." + "reason": "MobileNetV2 and DeBERTa v2 are unrelated code paths; not duplicates." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "Different model families (GPT-J/CodeGen vs MGP-STR). Same pattern, not the same change." + "reason": "GPT-J/codegen refactor vs MGP-STR refactor; same umbrella pattern only, not the same concrete change." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen and MPT are different implementations; not a duplicate." + "reason": "GPT-J/codegen and MPT touch different implementations; not the same bug or patch." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo output-tracing migration in the same model family and code path; these could be merged into one PR." + "reason": "Same GPT-Neo output-tracing refactor in the same file/test area; 44068 appears to be a later revision of the same change." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen vs GPT-Neo are different code paths, so this is not a duplicate pair." + "reason": "GPT-J/codegen vs GPT-Neo; same refactor style but unrelated code paths." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen are related but distinct model implementations." + "reason": "GPT-Neo and GPT-J are different model implementations, so this is not a duplicate." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet and GPT-J/CodeGen are unrelated model families; not the same underlying bug/change." + "reason": "ResNet output tracing and GPT-J/codegen output tracing are separate model-specific changes." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J/CodeGen are different code paths." + "reason": "MGP-STR and GPT-J/codegen are unrelated model families." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet vs GPT-J is not the same concrete change." + "reason": "ResNet and GPT-J are different code paths; same theme only." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet are separate model families and separate refactors." + "reason": "GPT-Neo vs ResNet; no shared underlying bug." }, { "left": "pull_request:44019", @@ -24193,319 +24256,241 @@ "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are different model code paths." + "reason": "GPT-Neo and GPT-J/codegen are separate model refactors." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert are different families; not the same bug/change." + "reason": "GPT-Neo and Flaubert/XLM are unrelated changes." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert are unrelated code paths." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44722", - "accept": false, - "reason": "MPT and GPT-J/CodeGen are separate implementations." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44085", - "accept": false, - "reason": "GPT-Neo and GPT-J are different model families; same refactor style is not enough." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-Neo vs MGP-STR are unrelated model code paths." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44085", - "accept": false, - "reason": "MPT and GPT-J are different implementations." - }, - { - "left": "pull_request:44068", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-Neo and MGP-STR are unrelated code paths." - }, - { - "left": "pull_request:44085", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-J and MGP-STR are different model families." + "reason": "GPT-Neo vs Flaubert/XLM; not duplicates." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT2 and MGP-STR are different code paths." + "reason": "GPT-2 and MGP-STR differ in model and code path." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert/XLM are different model families." + "reason": "ViLT and Flaubert/XLM are unrelated model-specific output-handling refactors." }, { "left": "pull_request:44059", "right": "pull_request:44722", "accept": false, - "reason": "GPT2 and GPT-J/CodeGen are different implementations." + "reason": "GPT-2 and GPT-J/codegen are separate model implementations; not mergeable as one PR." }, { - "left": "pull_request:44018", - "right": "pull_request:44071", + "left": "pull_request:44017", + "right": "pull_request:44161", "accept": false, - "reason": "GPT-Neo and MPT are unrelated code paths." + "reason": "SegFormer and LongT5 touch different subsystems and concrete changes." }, { "left": "pull_request:44073", - "right": "pull_request:44161", + "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and LongT5 are different model families." + "reason": "VisualBert and TextNet are unrelated model implementations." }, { - "left": "pull_request:44073", - "right": "pull_request:44101", + "left": "pull_request:44024", + "right": "pull_request:44076", "accept": false, - "reason": "VisualBert vs Flaubert/XLM is not the same concrete change." + "reason": "FocalNet and ImageGPT are different code paths; same refactor pattern only." }, { - "left": "pull_request:44059", - "right": "pull_request:44071", + "left": "pull_request:44073", + "right": "pull_request:44101", "accept": false, - "reason": "GPT2 and MPT are different model implementations." + "reason": "VisualBert vs Flaubert/XLM; no shared concrete bug or patch." }, { - "left": "pull_request:43997", - "right": "pull_request:44017", + "left": "pull_request:44056", + "right": "pull_request:44073", "accept": false, - "reason": "RegNet and SegFormer are different vision backbones and different output paths." + "reason": "MPNet and VisualBert are distinct model refactors." }, { - "left": "pull_request:43997", - "right": "pull_request:44056", + "left": "pull_request:44056", + "right": "pull_request:44071", "accept": false, - "reason": "RegNet and MPNet are unrelated model families." + "reason": "MPNet and MPT are unrelated implementations." }, { - "left": "pull_request:43997", - "right": "pull_request:44018", + "left": "pull_request:44071", + "right": "pull_request:44722", "accept": false, - "reason": "RegNet and GPT-Neo are unrelated code paths." + "reason": "MPT and GPT-J/codegen are different model code paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44161", + "left": "pull_request:44018", + "right": "pull_request:44059", "accept": false, - "reason": "RegNet and LongT5 are different implementations." + "reason": "GPT-Neo and GPT-2 are separate model-specific changes." }, { - "left": "pull_request:44074", - "right": "pull_request:44101", + "left": "pull_request:44059", + "right": "pull_request:44066", "accept": false, - "reason": "TextNet and Flaubert/XLM are different model families." + "reason": "GPT-2 vs GPT-J/codegen; same tracing theme, not the same bug." }, { - "left": "pull_request:44056", + "left": "pull_request:44101", "right": "pull_request:44161", "accept": false, - "reason": "MPNet and LongT5 are unrelated code paths." + "reason": "Flaubert/XLM and LongT5 are unrelated model implementations." }, { "left": "pull_request:43997", - "right": "pull_request:44074", - "accept": false, - "reason": "RegNet and TextNet are different backbones; same refactor theme is too broad." - }, - { - "left": "pull_request:44010", - "right": "pull_request:44071", + "right": "pull_request:44017", "accept": false, - "reason": "SqueezeBERT and MPT are different model families." + "reason": "RegNet and SegFormer are different model families." }, { - "left": "pull_request:44018", - "right": "pull_request:44129", + "left": "pull_request:43997", + "right": "pull_request:44056", "accept": false, - "reason": "GPT-Neo and SpeechT5 are different implementations." + "reason": "RegNet and MPNet are unrelated code paths." }, { "left": "pull_request:44071", "right": "pull_request:44076", "accept": false, - "reason": "MPT and ImageGPT are unrelated code paths." + "reason": "MPT and ImageGPT are distinct implementations; no overlap beyond the broad refactor theme." }, { "left": "pull_request:44074", - "right": "pull_request:44161", - "accept": false, - "reason": "TextNet and LongT5 are different model families." - }, - { - "left": "pull_request:44010", - "right": "pull_request:44073", + "right": "pull_request:44101", "accept": false, - "reason": "SqueezeBERT and VisualBert are unrelated model families." + "reason": "TextNet and Flaubert/XLM are unrelated model-specific changes." }, { - "left": "pull_request:44017", - "right": "pull_request:44056", + "left": "pull_request:44056", + "right": "pull_request:44161", "accept": false, - "reason": "SegFormer and MPNet are different implementations." + "reason": "MPNet and LongT5 are separate code paths." }, { - "left": "pull_request:44017", - "right": "pull_request:44074", + "left": "pull_request:43997", + "right": "pull_request:44018", "accept": false, - "reason": "SegFormer and TextNet are different vision/text model code paths." + "reason": "RegNet and GPT-Neo are unrelated model implementations." }, { - "left": "pull_request:44017", + "left": "pull_request:43997", "right": "pull_request:44161", "accept": false, - "reason": "SegFormer and LongT5 are unrelated model families." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44073", - "accept": false, - "reason": "SegFormer and VisualBert are different code paths." + "reason": "RegNet and LongT5 are different concrete changes." }, { - "left": "pull_request:44017", - "right": "pull_request:44722", + "left": "pull_request:44074", + "right": "pull_request:44161", "accept": false, - "reason": "SegFormer and GPT-J/CodeGen are unrelated implementations." + "reason": "TextNet and LongT5 are unrelated model refactors." }, { "left": "pull_request:44056", - "right": "pull_request:44101", - "accept": false, - "reason": "MPNet and Flaubert/XLM are different model families." - }, - { - "left": "pull_request:44073", - "right": "pull_request:44161", + "right": "pull_request:44129", "accept": false, - "reason": "VisualBert and LongT5 are unrelated code paths." + "reason": "MPNet and SpeechT5 are different code paths." }, { - "left": "pull_request:44073", - "right": "pull_request:44101", + "left": "pull_request:43997", + "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and Flaubert/XLM are different model families." + "reason": "RegNet and TextNet are separate model implementations." }, { - "left": "pull_request:44056", - "right": "pull_request:44074", + "left": "pull_request:44010", + "right": "pull_request:44018", "accept": false, - "reason": "MPNet and TextNet are unrelated implementations." + "reason": "SqueezeBERT and GPT-Neo are unrelated models; not duplicates." }, { "left": "pull_request:44071", - "right": "pull_request:44076", + "right": "pull_request:44129", "accept": false, - "reason": "MPT and ImageGPT are different model families." + "reason": "MPT and SpeechT5 are different model families." }, { "left": "pull_request:43996", - "right": "pull_request:44161", + "right": "pull_request:44017", "accept": false, - "reason": "FNet/CVT and LongT5 are unrelated code paths." + "reason": "FNet/CVT and SegFormer are unrelated changes." }, { "left": "pull_request:44010", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBERT and DeBERTaV2 are different model families." + "reason": "SqueezeBERT and DeBERTa v2 are distinct code paths." }, { "left": "pull_request:44044", - "right": "pull_request:44074", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa v2 and SpeechT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", "accept": false, - "reason": "DeBERTaV2 and TextNet are unrelated implementations." + "reason": "DeBERTa v2 and ViTS are different architectures and changes." }, { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "FocalNet and ImageGPT are different model families." + "reason": "FocalNet and ImageGPT are different model implementations." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and TextNet are different code paths." + "reason": "VisualBert and TextNet are unrelated code paths." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing migration on the same implementation; 44722 is effectively an expanded variant of the same change." + "reason": "Same GPT-J/codegen output-tracing work in the same files; 44722 looks like the later, broader version of 44066." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J output-tracing refactor in the same file family; these could be consolidated into one PR." + "reason": "Both are GPT-J output-tracing refactors on the same code path; they overlap enough to be treated as duplicate PRs." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J and SpeechT5 are different model families; not the same underlying change." + "reason": "GPT-J and SpeechT5 are unrelated models." }, { "left": "pull_request:43996", "right": "pull_request:44007", "accept": false, - "reason": "FNet/CVT and ResNet/RT-DETR ResNet are different model families and different code paths." + "reason": "FNet/CVT and RegNet are separate model refactors." }, { "left": "pull_request:43996", "right": "pull_request:44129", "accept": false, - "reason": "FNet/CVT and SpeechT5 are unrelated implementations." - }, - { - "left": "pull_request:44044", - "right": "pull_request:44129", - "accept": false, - "reason": "DeBERTaV2 and SpeechT5 are different model families." - }, - { - "left": "pull_request:44044", - "right": "pull_request:44154", - "accept": false, - "reason": "DeBERTaV2 and VITS are unrelated code paths." - }, - { - "left": "pull_request:44024", - "right": "pull_request:44076", - "accept": false, - "reason": "FocalNet and ImageGPT are different model families." - }, - { - "left": "pull_request:44073", - "right": "pull_request:44074", - "accept": false, - "reason": "VisualBert and TextNet are different code paths." + "reason": "FNet/CVT and SpeechT5 are unrelated code paths." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is supported by the packet: the cluster is a broad set of model-specific output-tracing refactors, and the accepted soft-edge pairs are limited to closely related GPT-J/CodeGen and GPT-Neo changes that look mergeable into single same-bugfix PRs. The rejected cross-model pairs are conservative and grounded in distinct model families/code paths." + "feedback": "The summary is grounded in the packet and stays conservative on the cross-model pairs. The accepted duplicate pairs are supported by shared files/targets and appear mergeable as single bugfix PRs, with 44722 a reasonable canonical representative of the GPT-J/codegen line." }, "evaluator_used": true, "retried": false, @@ -24539,14 +24524,14 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "b6e7538cf8726027249d12d25939921feeb4ed803b9f6e408d139403787d4019", + "prepared_review_unit_hash": "19898176a225f9d5e0f396d7554ae99c8303bc7c7c62ab2044f6fda829297702", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 76, - "serialized_chars": 84112, - "estimated_input_tokens": 21028, - "estimated_eval_tokens": 42312 + "serialized_chars": 84115, + "estimated_input_tokens": 21029, + "estimated_eval_tokens": 42314 }, "budget": { "node_count": 6, @@ -24561,36 +24546,36 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These PRs all implement the same broad refactor pattern\u2014migrating model forwards to standardized output tracing\u2014but they are model-specific changes, not duplicates of the same underlying fix. None of the soft edges should be merged as duplicates.", - "confidence": 0.9, + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they target different architectures and code paths, so they are not duplicates of one another.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue is present; as a representative PR, 44073 is a clear, self-contained example of the standardized output-capture refactor with explicit model-level tracing setup.", + "canonical_pr_reason": "PR 44129 is the strongest representative of the cluster because it has an explicit issue target and a focused, self-contained SpeechT5 output-capture refactor.", "best_issue_reason": null, - "best_pr_reason": "44073 is the most straightforward cluster anchor: it clearly introduces the capture_outputs/can_return_tuple refactor for a single model and reads like a canonical instance of the pattern.", + "best_pr_reason": "PR 44129 is the best standalone PR in this set: it is clearly scoped, tied to an issue, and the diff centers on one concrete model implementation.", "soft_edge_verdicts": [ { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Same refactor theme, but MPNet vs VisualBert are separate model implementations and the code changes are model-specific, so they are not the same underlying PR." + "reason": "Both are output-tracing refactors, but one is MPNet and the other is VisualBert; they touch different model implementations and are separate changes." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "Both adopt standardized output tracing, but one is MPNet and the other is TextNet; these are distinct model-specific migrations, not one mergeable duplicate change." + "reason": "These are separate model migrations (MPNet vs TextNet), not the same underlying bug or code change." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are unrelated model code paths; the similarity is only the shared decorator migration pattern, so this is not a duplicate change." + "reason": "RegNet and SqueezeBert are unrelated code paths; similarity is only at the level of a shared refactor pattern." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "Both touch output handling, but SqueezeBert and SpeechT5 are different models with different forward signatures and tracing needs, so they are not the same concrete fix." + "reason": "SqueezeBert and SpeechT5 modify different model files and different forward logic, so they should not be merged as duplicates." } ] }, @@ -24670,256 +24655,274 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "56f4f6fb19d870203767ff4aace1a411facb9e8a0322127118a8ae85cbcf546b", + "prepared_review_unit_hash": "5e554becee13f46801ec8a3122348e554c38c6b810fe9b7ebbdc326fc0a51037", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40169, - "estimated_input_tokens": 10043, - "estimated_eval_tokens": 20342 + "serialized_chars": 40179, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40169, - "estimated_input_tokens": 10043, - "estimated_eval_tokens": 20342 + "serialized_chars": 40179, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are all thematically similar output-tracing refactors, but they touch different model families and code paths. They are related by pattern, not by a shared underlying bug or a mergeable single change.", + "summary": "These PRs all follow the same broad \u201coutput tracing refactor\u201d theme, but they are model-specific changes in different code paths and should not be deduplicated as the same underlying change. No soft edge looks like a true duplicate/mergeable pair.", "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44003 is the strongest representative anchor: it is the most substantial output-tracing refactor here, adds recorder plumbing, and updates two closely related Mamba-family implementations.", + "canonical_pr_reason": "44003 is the broadest single refactor in the set, touching both `mamba` and `falcon_mamba` and introducing the tracing helper wiring; it is the best representative if a single PR must anchor the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 44003 is the best single representative of the cluster because it covers the core tracing refactor with the broadest concrete implementation rather than a narrower per-model adaptation.", + "best_pr_reason": "44003 is the strongest representative of this cluster because it contains the most substantial output-tracing implementation changes rather than just signature cleanup in one model file.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Same refactor pattern, but timm_backbone and upernet are different model code paths." + "reason": "Different model implementations (`timm_backbone` vs `upernet`); same refactor pattern, not the same code-path change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Different model families; only the refactor style is shared." + "reason": "Separate model-specific refactors (`timm_backbone` vs `vision_text_dual_encoder`), not one shared bug or mergeable change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Different architectures and implementations; not the same underlying change." + "reason": "`timm_backbone` and `univnet` are unrelated code paths; similarity is only at the refactor template level." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Backbone tracing cleanup vs Mamba output-recording refactor; different code paths." + "reason": "Different architectures and files; this is not the same underlying tracing bug/change." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Timm backbone and RWKV are unrelated model implementations." + "reason": "`timm_backbone` and `rwkv` are unrelated models, so this is not a duplicate fix." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "MobileNetV1 and UnivNet use different forward paths and output semantics." + "reason": "`mobilenet_v1` and `univnet` change different model-specific forward paths; not the same issue." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different model families; same output-tracing theme is not enough." + "reason": "Different subsystems (`mobilenet_v1` vs `upernet`); no shared concrete bug." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "MobileNetV1 and DepthAnything are unrelated concrete changes." + "reason": "`mobilenet_v1` and `depth_anything` are separate model refactors, not one mergeable change." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Different model wrappers and code paths; not one mergeable fix." + "reason": "The PRs target different encoder/decoder models and code paths; not duplicates." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "MobileNetV1 vs SpeechEncoderDecoder is only superficial similarity." + "reason": "`mobilenet_v1` and `speech_encoder_decoder` are unrelated implementations despite similar wording." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Vision-text dual encoder and UnivNet are different implementations, not the same bug." + "reason": "Different models (`vision_text_dual_encoder` vs `univnet`); same refactor style, not same change." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different model code paths; no shared concrete defect." + "reason": "These touch unrelated code paths and model families, so they are not duplicates." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Vision-text dual encoder refactor is unrelated to Mamba tracing plumbing." + "reason": "`vision_text_dual_encoder` and `mamba/falcon_mamba` address different model internals; not mergeable as one fix." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Different models and different output behavior; only the refactor motif matches." + "reason": "Different model families and different forward paths; no shared concrete bug." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "UnivNet and UperNet are unrelated changes to different forward methods." + "reason": "`univnet` and `upernet` are separate model implementations with independent refactors." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Different model families and code paths." + "reason": "The changes affect unrelated code paths, so this is not the same underlying change." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "UnivNet and DepthAnything do not share the same concrete fix." + "reason": "`univnet` and `depth_anything` are different model-specific tracing refactors, not a duplicate fix." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "UperNet and RWKV are unrelated implementations." + "reason": "`upernet` and `rwkv` are unrelated models; the similarity is only in the refactor pattern." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures; not a single mergeable change." + "reason": "Different architectures and files; no common concrete bug or code-path change." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "Mamba-family tracing refactor vs SuperPoint output cleanup; different code paths." + "reason": "`mamba/falcon_mamba` and `superpoint` are unrelated implementations; not the same underlying issue." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Mamba output tracing and CodeGen cache/output refactor are not the same underlying bug." + "reason": "`mamba/falcon_mamba` vs `codegen` is just shared refactor language, not the same fix." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Timm backbone and CodeGen are unrelated model implementations." + "reason": "`timm_backbone` and `codegen` are different code paths and cannot be merged as one PR." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families (`univnet` vs `codegen`); no shared concrete bug." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "`upernet` and `codegen` are unrelated model refactors, not duplicates." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "`vision_text_dual_encoder` and `codegen` touch different model internals; not the same change." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "DepthAnything and VisionEncoderDecoder are different model families." + "reason": "`depth_anything` and `vision_encoder_decoder` are separate model-specific refactors with no shared code-path bug." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "DepthAnything and SpeechEncoderDecoder are unrelated code paths." + "reason": "These affect different model families and are not the same underlying fix." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Different models; same tracing refactor theme only." + "reason": "`depth_anything` and `superpoint` are different models; shared theme is too broad for duplicate triage." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "DepthAnything and RWKV are different concrete changes." + "reason": "`depth_anything` and `rwkv` change unrelated forward paths; not duplicates." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "DepthAnything and DPR are unrelated implementations." + "reason": "`depth_anything` and `dpr` are unrelated model implementations." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "VisionEncoderDecoder and RWKV are different code paths." + "reason": "`vision_encoder_decoder` and `rwkv` do not share the same concrete bug or fix." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "Encoder-decoder wrappers for vision and speech are analogous, but not the same PR-worthy change." + "reason": "Different encoder-decoder stacks; same refactor wording, but not the same change." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "Different model code paths; only the decorator/refactor pattern matches." + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated code paths." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "VisionEncoderDecoder and DPR are unrelated implementations." + "reason": "`vision_encoder_decoder` and `dpr` are different model families, so this is not a duplicate." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "SpeechEncoderDecoder and SuperPoint are different concrete fixes." + "reason": "`speech_encoder_decoder` and `superpoint` are separate implementations; not mergeable as one PR." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "SpeechEncoderDecoder and RWKV are unrelated model paths." + "reason": "Different models and different tracing code paths; not the same bug." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Different architectures and output handling." + "reason": "`speech_encoder_decoder` and `dpr` are unrelated model changes." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "SuperPoint and RWKV are unrelated model implementations." + "reason": "`superpoint` and `rwkv` are different model internals; similarity is only superficial." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "SuperPoint and DPR do not fix the same underlying issue." + "reason": "These PRs touch unrelated models and should not be merged as duplicates." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "RWKV and DPR are different code paths and changes." + "reason": "`rwkv` and `dpr` are separate model-specific refactors, not the same underlying change." } ] }, @@ -24949,7 +24952,7 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "69a232c2e1bca9f57118b2fff935fd29c418558812c37d1044cdbbe5994253a5", + "prepared_review_unit_hash": "2bfc3e14e2fdd17df8894c45b3a7e98739c2872e1c05242c2270dbe1ab95c967", "component_budget": { "node_count": 2, "item_count": 2, @@ -24971,18 +24974,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate model-addition PRs (PenguinVL vs Voxtral TTS) that overlap only in shared Transformers scaffolding, not in the underlying code path or feature being added.", - "confidence": 0.98, + "summary": "These are two separate model-implementation PRs that only overlap on generic Transformers scaffolding (docs, auto-registry, package wiring). They do not appear to fix the same underlying bug or add the same model.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR #44662 is the better cluster representative: it is a fuller, non-draft implementation with substantial review activity and a broader set of model, processor, and test files.", "best_issue_reason": null, - "best_pr_reason": "Neither PR is a duplicate of the other; if forced to pick the stronger standalone PR, PenguinVL looks more mature and broadly reviewed, while Voxtral TTS is draft-level and issue-linked. But they should not be clustered as duplicates.", + "best_pr_reason": "PR #44662 is the strongest standalone PR to keep as reference because it is more complete and mature than the draft Voxtral TTS PR.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different model integrations for different architectures and use cases. The shared files are generic auto/docs registration plumbing, which is not enough to indicate the same underlying bug or change." + "reason": "Different model additions with different code paths: PenguinVL vs Voxtral TTS. The shared auto/docs files are generic integration points, not evidence of the same change." } ] }, @@ -25012,7 +25015,7 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "592e0af322ac0367c92d7f6a16512c4cc017d004c8cab2c902ab6e809fb11e59", + "prepared_review_unit_hash": "d14c82a8ca98e3980308e8538c247e04166c8c0fc173004c6744999b3dbf71e7", "component_budget": { "node_count": 2, "item_count": 2, @@ -25034,18 +25037,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs are related by the general theme of module/kernel fusion, but they are not the same underlying change: one adds a generic Module Fusion API, while the other implements n-to-1 fusion through KernelConfig and integration hooks.", - "confidence": 0.95, + "summary": "The two PRs are related around module/kernel fusion, but they are not the same change: #44979 adds the core Module Fusion API and tests, while #45363 builds a specific n-to-1 kernel-fusion path on top of KernelConfig with integrations and an example.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44979 is the broader foundational change: it introduces the generic Module Fusion API and new fusion abstractions, whereas PR 45363 is a separate, more specific KernelConfig-based implementation path.", + "canonical_pr_reason": "#44979 is the more fundamental, general-purpose API change: it introduces the module fusion abstraction itself and accompanying tests, making it the better canonical representative.", "best_issue_reason": null, - "best_pr_reason": "PR 44979 is the best representative of the cluster because it is the central API-level fusion feature rather than a downstream application of that idea.", + "best_pr_reason": "#44979 is the best representative of the cluster because it is the core infrastructure PR; #45363 is a narrower feature built on top of that abstraction.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Different concrete changes and code paths: generic module-fusion API vs. n-to-1 kernel fusion via KernelConfig; related area, not a duplicate PR." + "reason": "Related but distinct: one adds the fusion API primitives, the other implements kernel-config-driven n-to-1 fusion integration. They do not look like the same underlying code change and should not be merged as duplicates." } ] }, @@ -25075,12 +25078,12 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "98db58e7236caab0dd00d21e193fd0b100253607a29e29cfb677825be27e7bf7", + "prepared_review_unit_hash": "bc710f9be9e0e4a34a366571b66b7a64b5beaff3c156e7867c39907266d151ec", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -25088,7 +25091,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -25097,18 +25100,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests fix the same Mistral regex patch bug in the same code path (`_patch_mistral_regex`) against the same file and issue target; 45317 is the more complete version because it includes tests.", + "summary": "Both pull requests address the same AttributeError in `_patch_mistral_regex` by switching from `tokenizer.backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer` in the same code path. PR 45317 is the fuller version because it adds tests as well.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "45317 is the stronger canonical PR because it contains the same core fix as 45086 plus test coverage, making it the more complete representative of the duplicate pair.", + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same fix as 45086 plus test coverage, making it the more complete change set.", "best_issue_reason": null, - "best_pr_reason": "45317 is the best PR candidate because it directly addresses the bug, shares the same issue target, and adds regression tests in addition to the code fix.", + "best_pr_reason": "PR 45317 is the strongest representative: same bug fix, same file, same issue target, and additional validation via tests.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying AttributeError fix in `_patch_mistral_regex`, same target issue, same changed implementation path; 45317 simply adds tests to the same fix and could plausibly subsume 45086." + "reason": "Same concrete bug fix in the same function and file; 45317 is essentially 45086 plus tests, so they are duplicative and could be merged into one PR." } ] }, @@ -25138,12 +25141,12 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "b7b26d5f4e3752c755cef19578e32f13ebe79d404349b0df34b98bff06d81395", + "prepared_review_unit_hash": "40d3ff98d4cf267c16ccf2cb4a86d4ba842ea256f94421900a9518766237fac2", "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 }, @@ -25151,7 +25154,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6066, + "serialized_chars": 6065, "estimated_input_tokens": 1517, "estimated_eval_tokens": 3290 }, @@ -25160,18 +25163,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related to NemotronH model work, but they target different changes: one is a refactor/inheritance and conversion-mapping update, while the other introduces separate NemotronHDense/NemotronHSparse models plus docs, auto-mapping, and tests. They should not be treated as duplicates.", + "summary": "These PRs are related to Nemotron H model support, but they target different levels of the change: one is a narrower refactor/conversion-mapping update, the other is a broader model split into dense/sparse variants with docs and tests. They should not be clustered as duplicates.", "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the better canonical representative because it is the broader, more complete model-family change and includes the user-facing docs/tests and auto-mapping updates.", + "canonical_pr_reason": "pull_request:45569 is the broader, more complete implementation of the Nemotron H redesign, covering the dense/sparse model split plus docs/tests; pull_request:45115 is a narrower inheritance/conversion refactor.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 is the strongest overall candidate for the cluster since it covers the main architecture split and is more likely to subsume the surrounding NemotronH model work.", + "best_pr_reason": "pull_request:45569 is the best representative PR for this cluster because it contains the larger user-facing architectural change and supporting documentation/test updates.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Different underlying changes: #45115 is a refactor to inherit GraniteMoeHybrid and adjust conversion mappings, while #45569 adds dedicated NemotronHDense/NemotronHSparse implementations, docs, and tests. Shared files alone are not enough to merge them as one PR." + "reason": "Related Nemotron H work, but not the same concrete change: 45115 focuses on conversion/inheritance refactoring, while 45569 introduces separate dense/sparse model packages and docs/tests. They would not plausibly merge into one duplicate PR." } ] }, @@ -25201,12 +25204,12 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "1147195d3c0d767589ff8da7da89ba4b34aca56ac1e75296fcb00e2213369458", + "prepared_review_unit_hash": "3cde588e9554b04aa342e3f43da60395aa2cd3f650d7d3dced0e8ce694aa857d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6091, + "serialized_chars": 6090, "estimated_input_tokens": 1523, "estimated_eval_tokens": 3302 }, @@ -25214,7 +25217,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6091, + "serialized_chars": 6090, "estimated_input_tokens": 1523, "estimated_eval_tokens": 3302 }, @@ -25223,18 +25226,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two related PRs for heterogeneous model work, but not the same change: one adds per-layer configuration only, while the other includes configuration plus modeling/cache/masking support. Overlap is substantial, yet they are not strict duplicates.", - "confidence": 0.9, + "summary": "The two PRs are closely related but not duplicates: #45332 is the broader end-to-end heterogeneous model support, while #45333 is a narrower config-only subset. They share docs and configuration plumbing, but their scopes diverge materially in modeling/runtime changes.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45332 is the broader feature PR and subsumes the configuration work while also adding the modeling-side implementation, so it is the better canonical representative of the cluster.", + "canonical_pr_reason": "#45332 is the more representative PR for the cluster because it covers the full heterogeneous model feature set, including modeling, masking, cache utilities, integrations, docs, and tests, whereas #45333 only covers per-layer configuration support.", "best_issue_reason": null, - "best_pr_reason": "PR #45332 best represents the underlying effort because it covers the full heterogeneous-model feature set rather than only the configuration layer.", + "best_pr_reason": "#45332 is the best global PR to keep as the canonical representative because it implements the broader, more complete change set; #45333 is a partial follow-on focused on configuration only.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "They share the heterogeneous configuration docs and core config files, but #45332 extends into modeling, masking, cache, and skip utilities while #45333 is limited to configuration support. That looks like a subset/dependent PR, not the same concrete change that could simply be merged as one duplicate PR." + "reason": "They overlap on heterogeneous configuration/docs, but #45332 also changes modeling, masking, cache, and integration code paths. #45333 is config-only, so these are related changes rather than the same underlying PR." } ] }, @@ -25264,40 +25267,40 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "5ea57ae5a48d6fd6a9d66f2e1b315733ebf6e19bdaf12279547041b322cbeab2", + "prepared_review_unit_hash": "07c765dcaf16f0a1eba7a72926de65ab4447c1761d1d1eed6d6bb85dde8eda6f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5512, - "estimated_input_tokens": 1378, - "estimated_eval_tokens": 3012 + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5512, - "estimated_input_tokens": 1378, - "estimated_eval_tokens": 3012 + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are typing-related, but they address different changes: #45415 is a broad repo-wide type-checking pass across many `src/transformers/*` files, while #45425 adds shared typing helpers/protocols and related fixes centered on `_typing.py`, `modeling_utils`, and a few integrations. They are not the same underlying change.", - "confidence": 0.93, + "summary": "These are related typing-related PRs, but they are not the same underlying change: one is a broad type-checking sweep across many transformers modules, while the other adds targeted typing helper protocols and compatibility shims.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the two PRs overlap on typing infrastructure but target different code changes and would not plausibly be merged as one duplicate fix.", + "canonical_pr_reason": "PR 45415 is the broader, more representative typing change: it updates type-checking config and touches many runtime modules in one coherent sweep.", "best_issue_reason": null, - "best_pr_reason": "#45425 is the more self-contained representative of the typing work because it introduces reusable typing protocol definitions and related fixes, whereas #45415 is a much broader type-checking sweep.", + "best_pr_reason": "PR 45415 is the best single representative of the cluster because its scope and intent are clearer and more central than the narrower helper-typing update in 45425.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Shared typing theme and a few filenames, but different concrete scope and purpose: one is a mass type-checking cleanup, the other adds typing protocols/helpers and targeted fixes. Not the same underlying change." + "reason": "Do not merge: 45415 is a wide type-checking pass across many files, while 45425 adds specific protocol/type-alias helpers in _typing.py and related integrations. They share a typing theme and a few files, but not the same concrete bug/change or mergeable PR scope." } ] }, @@ -25327,12 +25330,12 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "6e8726ae5e98f927e0308f387faf72fd8402133cfa90fe5510c7c39c5234b3a4", + "prepared_review_unit_hash": "f59795662437af70fa5efd159cc666810c793231efad818b41fc6c27193bbb85", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4507, + "serialized_chars": 4508, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -25340,7 +25343,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4507, + "serialized_chars": 4508, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -25349,18 +25352,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related to expert-parallel loading, but they fix different bugs in different code paths/backends: one is about EP + DeepSpeed ZeRO-3 config/launch handling, the other about EP + FSDP2 expert parameters being overwritten during loading. Not a duplicate pair.", - "confidence": 0.93, + "summary": "Both PRs are about expert-parallel / distributed loading behavior, but they address different failure modes: one fixes serialization and ZeRO-3 loading metadata, the other fixes shard wrapping to prevent expert weights from being overwritten under FSDP2/broadcast. They are related in subsystem, not the same underlying bug or change.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45662 is the better representative of the cluster because it targets a concrete parameter-loading corruption path with explicit model-loading and sharding logic changes, whereas 45548 is a narrower accelerate/ZeRO-3 configuration-loading fix.", + "canonical_pr_reason": "No clear duplicate canonical PR: these PRs fix different distributed-loading bugs and should remain separate.", "best_issue_reason": null, - "best_pr_reason": "PR 45662 is the strongest standalone candidate: it fixes a specific, high-impact EP/FSDP2 bug where experts can be silently overwritten during load, with code changes directly addressing the affected path.", + "best_pr_reason": "PR #45662 is the stronger representative of the cluster because it targets a concrete runtime corruption bug in the loading path and has broader code-path coverage, though it is not a duplicate of #45548.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Both mention EP and loading, but the failures differ: 45548 is about DeepSpeed ZeRO-3/accelerate-launch and serializing distributed_config, while 45662 is about FSDP2/rank-0 broadcast overwriting expert weights. Different concrete bugs and not safely mergeable into one PR." + "reason": "Reject: both involve EP/distributed loading, but #45548 fixes config serialization and ZeRO-3/accelerate launch loading, while #45662 fixes post-shard wrapping to stop experts being overwritten under FSDP2. Different concrete bugs and patches." } ] }, @@ -25390,40 +25393,40 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "d14b567117447ea46e4916d308a1d6565632ec8b519f0458eef8b42de82769fd", + "prepared_review_unit_hash": "bd9d81806ac3fc92dce5d027a992ac788d4dd4790e7bab98330b4d1c5b856f06", "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 address the same xdist-safe captured_info logging problem in transformers testing utilities. PR 45645 appears to subsume 45639 by adding the same worker-safe output-file change plus CI workflow updates to collect multiple captured_info files, so they are duplicate/overlapping change candidates.", - "confidence": 0.94, + "summary": "Both PRs are about making captured debug output from patched tests safe under pytest-xdist and easier to collect in CI, but they are not the same concrete change: one focuses on the logging helper, while the other also changes the workflow that aggregates and prints the artifacts.", + "confidence": 0.78, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the more complete representative: it contains the xdist-safe testing_utils fix and also updates CI log collection to handle multiple captured_info files, covering the broader end-to-end change.", + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the broader end-to-end fix: xdist-safe artifact writing plus CI-side collection/printing of multiple captured_info files.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 best represents the cluster because it includes the core bug fix and the downstream workflow adjustment needed for the new file naming scheme; 45639 is a narrower subset of the same change.", + "best_pr_reason": "PR 45645 is the best representative of the cluster since it includes the core helper change and the downstream workflow handling needed to preserve the debug logs in CI.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", - "accept": true, - "reason": "Same underlying bug: xdist collisions for captured_info output. The second PR extends the first with compatible workflow handling, and both could plausibly be merged as one change." + "accept": false, + "reason": "Related, but not the same PR change: 45639 only hardens the testing helper\u2019s file output, while 45645 additionally changes the CI workflow to gather and print multiple captured_info files. That extra workflow behavior makes them different concrete fixes rather than a single mergeable PR." } ] }, @@ -25519,40 +25522,40 @@ "soft_pairs": [ "pull_request:45932|pull_request:46100" ], - "prepared_review_unit_hash": "cc94ecc36c4d189990d857999946a81163151d4c5be55f513aa2a3e25b9d19f9", + "prepared_review_unit_hash": "746c0332085adb1e52ce51a6b661afde007397b60fd161fba02feee1c8aebf1e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5161, - "estimated_input_tokens": 1291, - "estimated_eval_tokens": 2838 + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5161, - "estimated_input_tokens": 1291, - "estimated_eval_tokens": 2838 + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs are not duplicates: one is a CI/telemetry and model-config related change, while the other is a test migration adding DDP/kernel test coverage. They share only broad testing/infra context, not the same concrete bug or code path.", - "confidence": 0.93, + "summary": "These two PRs are not duplicates: one mixes CI/telemetry changes with model config adjustments, while the other is a test-migration PR adding DDP/kernel test infrastructure. They share only broad \"do not merge\"/testing vibes, not the same underlying code-path change.", + "confidence": 0.91, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #46100 is the more self-contained and concrete change set; PR #45932 is a broader, unrelated testing/CI/workflow mix and is a poor representative for duplicate clustering.", + "best_pr_reason": "Neither PR is a good canonical duplicate target; they address different concerns and should remain separate. If forced, 46100 is the more self-contained change, but it is not a duplicate of 45932.", "soft_edge_verdicts": [ { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "Reject: the diffs target different problems and code paths. #45932 changes CI secrets/workflows and several model files, while #46100 adds kernel-test migration files and vit-related tests. Shared testing context is too broad to treat as the same change." + "reason": "Different changesets and objectives: 45932 touches CI secrets plus model embedding/config code, while 46100 is a test migration adding DDP config and new kernel/ViT tests. No shared concrete bug or mergeable combined fix." } ] }, @@ -25582,7 +25585,7 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "f810362454c97ee363a1622aad19a57ead811fc3adc217c4d7a9ee0abfe0b1de", + "prepared_review_unit_hash": "8754d528c860337f1224d840dcec7e19c2e253452f71863df40305a3c6eedd89", "component_budget": { "node_count": 2, "item_count": 2, @@ -25604,18 +25607,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs overlap in a few distributed filenames, but they are not the same concrete change: #46126 is a focused EP/trainer integration patch, while #46269 is a broad distributed branch base touching many unrelated distributed areas. This is not a safe duplicate/merge candidate pair.", - "confidence": 0.91, + "summary": "The pair is not a true duplicate: PR #46126 is a focused EP/trainer integration change, while PR #46269 is a broad distributed branch base with many unrelated distributed/FSDP/CI edits. The shared filenames are incidental overlap, not the same underlying fix or feature.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46126 is the better canonical representative because it is the narrowly-scoped, concrete EP + Trainer integration change with a clear code-path purpose.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #46126 is the best fit overall: it targets one specific distributed/MoE behavior, whereas #46269 is an umbrella base branch spanning many features and subsystems.", + "best_pr_reason": "PR #46126 is the more specific, coherent change centered on EP routing and Trainer integration; PR #46269 is too broad and mixed to serve as a canonical duplicate representative.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Reject: shared distributed filenames are not enough here. #46126 implements a specific EP routing/trainer integration change, while #46269 is a broad distributed branch base touching many areas and is not plausibly a single merged fix with #46126." + "reason": "Different scope and purpose: #46126 adds router-parallel/Trainer integration, while #46269 is a large distributed base branch spanning FSDP, CI, and many unrelated files. They do not look like the same concrete code-path fix." } ] }, @@ -25683,12 +25686,12 @@ "pull_request:46504|pull_request:46716", "pull_request:46432|pull_request:46716" ], - "prepared_review_unit_hash": "039c22a6ed1349477446f7d3f0ee9443b70764e38852a0308be8d0b405eaec2c", + "prepared_review_unit_hash": "22e3bbdcca07d1baaafd975ba2f030febcb738a8c418209799ad9596cf8bfb51", "component_budget": { "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26678, + "serialized_chars": 26679, "estimated_input_tokens": 6670, "estimated_eval_tokens": 13596 }, @@ -25696,7 +25699,7 @@ "node_count": 10, "item_count": 10, "soft_pair_count": 31, - "serialized_chars": 26678, + "serialized_chars": 26679, "estimated_input_tokens": 6670, "estimated_eval_tokens": 13596 }, @@ -25705,198 +25708,198 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated workflow-chore PRs, plus a small family of similarly titled pr-ci-security-gate changes. I found no true duplicate pairs; the soft similarities are mostly generic CI/security hardening overlaps rather than the same concrete fix.", - "confidence": 0.82, + "summary": "All items are small, unrelated workflow/security chore PRs. The only dense similarity cluster is the pr-ci-security-gate.yml group, but those PRs make different changes and should not be merged as duplicates.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46554 is the most substantive and representative change in the security-gate mini-cluster: it adds the actual checkout/diff-based check logic, whereas nearby PRs are follow-up permission/secrets tweaks or comment-only updates.", + "canonical_pr_reason": "No true duplicate PR stands out; 46554 is the most substantive and central workflow-security change, so it is the best representative if one must be chosen.", "best_issue_reason": null, - "best_pr_reason": "PR 46554 is the best standalone PR in the set because it implements a concrete security check with clear functional impact; the other PRs are narrower workflow adjustments or unrelated CI chores.", + "best_pr_reason": "46554 has the broadest security-gate logic change and is the most representative anchor for the workflow-security subgroup, but it is still distinct from the nearby PRs.", "soft_edge_verdicts": [ { "left": "pull_request:46138", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different concerns: stdin handling in self-comment CI vs PR security-gate logic." + "reason": "Different workflow files and different changes: stdin piping for self-comment CI vs PR security-gate logic." }, { "left": "pull_request:46138", "right": "pull_request:46558", "accept": false, - "reason": "Different workflows and different concerns: stdin handling in self-comment CI vs security-gate permissions/approval changes." + "reason": "Different workflows and different code paths; no shared underlying bug/change." }, { "left": "pull_request:46138", "right": "pull_request:46563", "accept": false, - "reason": "Comment-only reusable-workflow note vs stdin input handling; not the same fix." + "reason": "Self-comment CI message handling is unrelated to security-gate workflow notes." }, { "left": "pull_request:46138", "right": "pull_request:46716", "accept": false, - "reason": "Secrets-handling tweak in the PR security gate vs stdin input handling in self-comment CI; unrelated." + "reason": "Unrelated workflow updates in different files; not the same fix." }, { "left": "pull_request:46138", "right": "pull_request:46504", "accept": false, - "reason": "Different workflows: PR comment stdin handling vs commit-message parsing for build-image tagging." + "reason": "Different workflows: message input handling vs build-ci-docker image tag detection." }, { "left": "pull_request:46432", "right": "pull_request:46503", "accept": false, - "reason": "SSH runner hardening vs benchmark trigger condition; different workflows and bugs." + "reason": "SSH runner hardening is unrelated to benchmark_v2 dispatch logic." }, { "left": "pull_request:46432", "right": "pull_request:46501", "accept": false, - "reason": "SSH runner hardening vs anti-slop workflow edit; unrelated." + "reason": "Different workflows and concerns; SSH runner vs anti-slop permissions." }, { "left": "pull_request:46432", "right": "pull_request:46502", "accept": false, - "reason": "SSH runner hardening vs reviewer assignment workflow; unrelated." + "reason": "SSH runner changes do not match assign-reviewers checkout behavior." }, { "left": "pull_request:46432", "right": "pull_request:46554", "accept": false, - "reason": "SSH-runner security tweaks vs PR security-gate logic; different code paths." + "reason": "Different workflows and different security concerns; not one concrete bug." }, { "left": "pull_request:46432", "right": "pull_request:46558", "accept": false, - "reason": "SSH-runner security tweaks vs PR security-gate permissions/approval changes; different code paths." + "reason": "SSH runner env cleanup is unrelated to PR CI security-gate permissions/approval logic." }, { "left": "pull_request:46501", "right": "pull_request:46503", "accept": false, - "reason": "Anti-slop workflow edit vs benchmark trigger edit; unrelated." + "reason": "Different workflow jobs with different behaviors; only broad CI similarity." }, { "left": "pull_request:46501", "right": "pull_request:46502", "accept": false, - "reason": "Anti-slop workflow edit vs reviewer assignment workflow; unrelated." + "reason": "Anti-slop permissions and assign-reviewers checkout are distinct workflow changes." }, { "left": "pull_request:46501", "right": "pull_request:46504", "accept": false, - "reason": "Anti-slop workflow edit vs build-ci-docker-images commit-message logic; unrelated." + "reason": "Different workflows and fixes; anti-slop vs build image tagging." }, { "left": "pull_request:46501", "right": "pull_request:46554", "accept": false, - "reason": "Anti-slop workflow edit vs PR security-gate logic; different concerns." + "reason": "Anti-slop workflow permissions change is unrelated to PR CI security gate changes." }, { "left": "pull_request:46502", "right": "pull_request:46503", "accept": false, - "reason": "Reviewer assignment workflow vs benchmark trigger workflow; unrelated." + "reason": "Assign-reviewers base-ref checkout and benchmark dispatch are unrelated changes." }, { "left": "pull_request:46502", "right": "pull_request:46504", "accept": false, - "reason": "Reviewer assignment workflow vs build-image tag logic; unrelated." + "reason": "Different workflow files and different code paths." }, { "left": "pull_request:46502", "right": "pull_request:46554", "accept": false, - "reason": "Reviewer assignment workflow vs PR security-gate logic; unrelated." + "reason": "Assign-reviewers checkout fix is unrelated to PR CI security-gate logic." }, { "left": "pull_request:46503", "right": "pull_request:46504", "accept": false, - "reason": "Benchmark scheduling/dispatch vs build-image tag detection; different workflows and fixes." + "reason": "Benchmark dispatch support and build image tagging are different changes." }, { "left": "pull_request:46503", "right": "pull_request:46554", "accept": false, - "reason": "Benchmark trigger edit vs PR security-gate logic; unrelated." + "reason": "Benchmark_v2 trigger update is unrelated to PR CI security gate behavior." }, { "left": "pull_request:46504", "right": "pull_request:46554", "accept": false, - "reason": "Build-image tag parsing vs PR security-gate checks; different workflows and different problems." + "reason": "Build-ci-docker input quoting and PR security-gate changes are distinct." }, { "left": "pull_request:46504", "right": "pull_request:46558", "accept": false, - "reason": "Build-image tag parsing vs security-gate permissions/approval changes; unrelated." + "reason": "Different workflow purpose: build image tag selection vs security-gate permissions." }, { "left": "pull_request:46554", "right": "pull_request:46558", "accept": false, - "reason": "Same workflow file, but 46554 adds checkout/diff-based security checks while 46558 only adjusts permissions and approval gating; related, not the same fix." + "reason": "Same file, but different security-gate edits affecting different parts of the workflow." }, { "left": "pull_request:46554", "right": "pull_request:46563", "accept": false, - "reason": "46563 is a comment-only note about reusable-workflow safety, not the same functional change as 46554." + "reason": "Different changes in the same workflow file; not the same concrete fix." }, { "left": "pull_request:46554", "right": "pull_request:46716", "accept": false, - "reason": "46554 changes the gate logic; 46716 only removes inherited secrets from the reusable workflow call." + "reason": "Both touch the security-gate file, but they change different trust/permission settings." }, { "left": "pull_request:46558", "right": "pull_request:46563", "accept": false, - "reason": "Permissions/approval tweak vs comment-only note; not the same bug or change." + "reason": "Same workflow file, but permissions/approval logic vs review-comment notes are distinct." }, { "left": "pull_request:46558", "right": "pull_request:46716", "accept": false, - "reason": "Permissions/approval tweak vs secrets removal; related hardening, but different concrete changes." + "reason": "Distinct edits to the same workflow; not a duplicate change." }, { "left": "pull_request:46563", "right": "pull_request:46716", "accept": false, - "reason": "Comment-only note vs secrets inheritance change; not the same fix." + "reason": "Commentary about reusable workflow safety vs secrets handling are different changes." }, { "left": "pull_request:46504", "right": "pull_request:46563", "accept": false, - "reason": "Build-ci-docker-images commit parsing vs reusable-workflow safety note; unrelated." + "reason": "Build-ci-docker tagging logic is unrelated to security-gate documentation/comments." }, { "left": "pull_request:46432", "right": "pull_request:46563", "accept": false, - "reason": "SSH-runner hardening vs reusable-workflow safety note; unrelated." + "reason": "SSH runner hardening is unrelated to security-gate comments." }, { "left": "pull_request:46504", "right": "pull_request:46716", "accept": false, - "reason": "Build-image commit parsing vs secrets inheritance change in the PR security gate; unrelated." + "reason": "Build image tag logic and PR CI secrets handling are unrelated." }, { "left": "pull_request:46432", "right": "pull_request:46716", "accept": false, - "reason": "SSH-runner hardening vs secrets inheritance removal in the PR security gate; unrelated." + "reason": "SSH runner setup changes do not match PR CI secrets handling." } ] }, @@ -25989,7 +25992,7 @@ "soft_pairs": [ "pull_request:46281|pull_request:46890" ], - "prepared_review_unit_hash": "2dbed949d3ad5689a3cbeb523c2a3b49bd43fff3dccac956e3da75218434cb1d", + "prepared_review_unit_hash": "18b0289cc6228dd3645d0530cd09b318137905973fa4f6523890424071daea2f", "component_budget": { "node_count": 2, "item_count": 2, @@ -26011,18 +26014,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate security-hardening PRs for different model-loading paths: one blocks path traversal in sharded checkpoint filenames from `weight_map`, the other blocks traversal in the `transformers_weights` config field. They share a theme, but not the same underlying bug or mergeable change.", - "confidence": 0.92, + "summary": "These are two separate security hardening PRs that both address path traversal, but in different checkpoint-loading paths and with different inputs. They are related thematically, not duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: 46281 hardens sharded checkpoint filename resolution in `hub.py`, while 46890 validates `transformers_weights` in `modeling_utils.py`; they target different code paths and different attacker-controlled fields.", "best_issue_reason": null, - "best_pr_reason": "No single canonical PR: PR #46281 fixes sharded-checkpoint shard file resolution in `src/transformers/utils/hub.py`, while PR #46890 validates an explicit `transformers_weights` path in `src/transformers/modeling_utils.py`. Different code paths, different inputs, different fixes.", + "best_pr_reason": "No single best PR for deduping; both are legitimate but distinct fixes, so neither should represent the other.", "soft_edge_verdicts": [ { "left": "pull_request:46281", "right": "pull_request:46890", "accept": false, - "reason": "Both are path-traversal mitigations, but they target different loaders and different metadata fields (`weight_map` shard names vs. `transformers_weights`). The diffs do not indicate the same concrete bug or a PR that could reasonably be merged as one change." + "reason": "Different underlying bugs and code paths: one validates shard files from a weight map in `src/transformers/utils/hub.py`, the other blocks traversal via `transformers_weights` in `src/transformers/modeling_utils.py`. They would not plausibly merge into one PR." } ] }, @@ -26052,12 +26055,12 @@ "soft_pairs": [ "pull_request:46574|pull_request:46580" ], - "prepared_review_unit_hash": "645d93ce3d204fa13c4ff2cbc277ca41fde505c8aff44252432dfff1b0e4dffc", + "prepared_review_unit_hash": "385578fbe1919fad756b03f5af5c3d712d2f18c6e5183534cc4f60a73f6caf21", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4409, + "serialized_chars": 4410, "estimated_input_tokens": 1103, "estimated_eval_tokens": 2462 }, @@ -26065,7 +26068,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4409, + "serialized_chars": 4410, "estimated_input_tokens": 1103, "estimated_eval_tokens": 2462 }, @@ -26074,18 +26077,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs address unrelated tokenizer bugs in different code paths: CodeLlama decode-space stripping vs. AutoTokenizer selecting the wrong class for certain DeepSeek checkpoints. They should not be clustered as duplicates.", - "confidence": 0.96, + "summary": "These are unrelated pull requests: one fixes CodeLlama decode-time space stripping, the other fixes DeepSeek checkpoints loading under the wrong tokenizer class.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46580 is the better canonical representative for this cluster because it targets the central AutoTokenizer resolution logic and fixes a clearer class-selection bug with explicit checkpoint-prefix handling.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #46580 is the strongest standalone PR in the set: it has a well-scoped failure mode, explicit checkpoint targeting, and a broader user impact than the CodeLlama decode fix.", + "best_pr_reason": "pull_request:46574 is the stronger standalone bug fix: it targets one concrete tokenizer decode path and includes a direct issue link and test coverage. pull_request:46580 is also valid, but it is a separate AutoTokenizer routing fix for different checkpoints.", "soft_edge_verdicts": [ { "left": "pull_request:46574", "right": "pull_request:46580", "accept": false, - "reason": "Different underlying problems and code paths: one changes CodeLlama decoding behavior, the other changes AutoTokenizer class selection for DeepSeek checkpoints. They would not plausibly be merged as one PR." + "reason": "Different underlying problems and code paths: CodeLlama decode stripping vs DeepSeek checkpoint-to-tokenizer-class selection. They are not mergeable as one PR." } ] }, @@ -26117,7 +26120,7 @@ "pull_request:46781|pull_request:46993", "pull_request:46993|pull_request:46994" ], - "prepared_review_unit_hash": "c9a1ec26cc91ef350489f1e2d2578c1ad1bdd3341c845f07f45cb73791e3526e", + "prepared_review_unit_hash": "4410e1abf58e9736861a2fce68a41e1265d364f7ddabfe4fdf9f3a094ed3f076", "component_budget": { "node_count": 3, "item_count": 3, @@ -26139,24 +26142,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three pull requests are only loosely related through generation/cache code, but they fix different bugs: cache-layer validation, prompt lookup length capping, and empty-list validation in generation logits processing. No duplicate cluster is present.", - "confidence": 0.97, + "summary": "All three pull requests are separate fixes in adjacent generation/cache validation areas, but they address different code paths and failure modes, so they should not be clustered as duplicates.", + "confidence": 0.95, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 46993 is the strongest standalone bugfix in the set: it corrects prompt lookup decoding exceeding max_length with a targeted code-path change and test coverage.", + "best_pr_reason": "PR #46994 is the clearest standalone fix in the set: it adds explicit validation for empty entries in `bad_words_ids` / `sequence_bias` and tightens the corresponding error message. The other PRs solve unrelated bugs in prompt lookup decoding and QuantizedCache layer support.", "soft_edge_verdicts": [ { "left": "pull_request:46781", "right": "pull_request:46993", "accept": false, - "reason": "Different problems in different code paths: QuantizedCache layer-type validation vs prompt lookup decoding length capping. Shared generation/cache context is too broad to treat as the same change." + "reason": "Different subsystems and bugs: one is early rejection of unsupported QuantizedCache layer types, the other is prompt lookup decoding exceeding `max_length`. They do not target the same underlying behavior." }, { "left": "pull_request:46993", "right": "pull_request:46994", "accept": false, - "reason": "Both are generation-related, but one fixes prompt lookup max_length behavior while the other adds validation for empty token lists in bad_words_ids/sequence_bias. They are not the same underlying bug or mergeable into one fix." + "reason": "Both are generation-related, but they fix different problems: prompt lookup candidate length calculation vs validation of empty token lists in logits processors. They would not plausibly be merged as one PR." } ] }, @@ -26189,52 +26192,52 @@ "pull_request:47016|pull_request:47066", "pull_request:47048|pull_request:47066" ], - "prepared_review_unit_hash": "2c30cab3009395a8f0e48631e954b6d2bff4150cb2ad3ac8851a7dcde1f1c4ac", + "prepared_review_unit_hash": "a4232945901083c549cd2fed0926b1e54b41b28068d7476c3f08919efec65eae", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7336, - "estimated_input_tokens": 1834, - "estimated_eval_tokens": 3924 + "serialized_chars": 7337, + "estimated_input_tokens": 1835, + "estimated_eval_tokens": 3926 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7336, - "estimated_input_tokens": 1834, - "estimated_eval_tokens": 3924 + "serialized_chars": 7337, + "estimated_input_tokens": 1835, + "estimated_eval_tokens": 3926 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs all address the same failure mode: SciPy Hungarian matching can crash when the cost matrix contains NaN/Inf, usually from AMP/fp16 overflow. PR 47048 and 47066 are close duplicates at the bug-fix level across different object-detection model matchers; PR 47016 is the same symptom but in a separate RT-DETR code path.", - "confidence": 0.87, + "summary": "All three PRs address the same class of failure: Hungarian matching can crash when the cost matrix contains NaN/Inf, and each PR adds a nan_to_num guard in model-specific object-detection loss code. PR 47048 is the best representative because it is the broadest of the set and most centrally covers the issue pattern.", + "confidence": 0.86, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47048 is a strong representative of the cluster because it fixes the non-finite cost-matrix crash in two matcher implementations with a clear, shared guard and tests, making it a good anchor for the broader bug class.", + "canonical_pr_reason": "PR 47048 is the strongest canonical representative: it fixes the exact crash mode with a minimal guard in the shared DETR-style matcher path and covers two model families, making it a good anchor for the duplicate cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 47048 is the best single PR to keep because it directly patches the Hungarian matcher failure in a reusable object-detection loss path and is broad enough to represent the underlying defect without overfitting to one model.", + "best_pr_reason": "PR 47048 is the best single PR to represent the cluster because it is broad, directly targeted at the non-finite cost-matrix crash, and has the clearest generalizable fix pattern among the three.", "soft_edge_verdicts": [ { "left": "pull_request:47016", "right": "pull_request:47048", "accept": false, - "reason": "Same symptom, but different model-specific matcher paths (RT-DETR vs DETR/Grounding DINO). They are related but not the same concrete code path, so I would not dedupe them." + "reason": "Same symptom, but different model-specific code paths (RT-DETR vs DETR/Grounding DINO). This looks like parallel hardening of related matchers rather than the same concrete change that should be merged as one PR." }, { "left": "pull_request:47016", "right": "pull_request:47066", "accept": false, - "reason": "Both guard against infeasible cost matrices, but they patch different matcher implementations in different model families. Too broad to treat as the same PR." + "reason": "Both fix Hungarian matcher infeasibility, but they touch different model families and different loss implementations. That is too broad to treat as one duplicate PR pair under a strict code-path standard." }, { "left": "pull_request:47048", "right": "pull_request:47066", "accept": true, - "reason": "Both PRs fix the same concrete bug class in Hungarian matching: non-finite cost matrices causing scipy linear_sum_assignment to fail. The changes are highly analogous and could plausibly be merged into one broader PR." + "reason": "These two PRs make the same concrete fix pattern in the same kind of Hungarian-matching loss code: guard non-finite costs before scipy assignment. They are close enough in mechanism and scope to plausibly be merged into one broader PR." } ] }, @@ -26264,7 +26267,7 @@ "soft_pairs": [ "pull_request:47026|pull_request:47046" ], - "prepared_review_unit_hash": "33c160948750802429a005a4d7e8fc1aef7c7ec8dd3a9f28fd0a4dad3d197c90", + "prepared_review_unit_hash": "fca1e7916fb4c3b26c55bcc162ae2ee4d2d8d21ae5949053462667a5c29a16e9", "component_budget": { "node_count": 2, "item_count": 2, @@ -26286,18 +26289,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs target the same xLSTM forward-path crash family in training mode around `return_last_states`/`use_cache` handling in `modeling_xlstm.py`. PR 47046 looks like the broader, more complete follow-up, while 47026 is the narrower earlier fix.", - "confidence": 0.82, + "summary": "Both pull requests appear to address the same xLSTM train-mode crash around `return_last_states=False` in `modeling_xlstm.py`; the second PR looks like a broader, more complete fix with tests and an additional `use_cache` mention.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47046 is the better canonical PR because it addresses the same xLSTM train-mode crash path but with a broader fix surface (`return_last_states=False` and `use_cache`) and adds tests, making it the more complete representative.", + "canonical_pr_reason": "PR 47046 is the better canonical PR: it targets the same xLSTM forward-path crash, adds tests, and appears to subsume the narrower fix in 47026.", "best_issue_reason": null, - "best_pr_reason": "PR 47046 is the best overall representative of the cluster since it covers the core xLSTM train-mode failure and extends the fix beyond the narrower crash in 47026.", + "best_pr_reason": "PR 47046 is the strongest representative because it covers the same bug and looks more complete/validated than 47026.", "soft_edge_verdicts": [ { "left": "pull_request:47026", "right": "pull_request:47046", "accept": true, - "reason": "Same xLSTM `modeling_xlstm.py` train-mode crash area, same issue target, and overlapping `return_last_states=False` forward-path fix; 47046 appears to be a broader version of the same underlying bug rather than a different change." + "reason": "Same underlying xLSTM train-mode crash path in the same file, with the same issue target; 47046 looks like the more complete version of the fix." } ] }, @@ -26339,12 +26342,12 @@ "pull_request:47060|pull_request:47076", "pull_request:47074|pull_request:47076" ], - "prepared_review_unit_hash": "1b146a42b690a13e4f491e6ba45f156c18e1b8bf26876983fb294777d5d7a17b", + "prepared_review_unit_hash": "c19d4d7da4c86054340120ffba43860549fbeeec21a3fa506d13c3fc33215a9b", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 9, - "serialized_chars": 15762, + "serialized_chars": 15764, "estimated_input_tokens": 3941, "estimated_eval_tokens": 8138 }, @@ -26352,7 +26355,7 @@ "node_count": 6, "item_count": 6, "soft_pair_count": 9, - "serialized_chars": 15762, + "serialized_chars": 15764, "estimated_input_tokens": 3941, "estimated_eval_tokens": 8138 }, @@ -26361,66 +26364,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are all multimodal/pipeline-processing related, but they address different concrete bugs: stop-sequence forwarding, audio format validation, generation-kwarg routing, batched decoding, input-key normalization, and Gemma4 placeholder checks. They share subsystem overlap, not a single underlying change.", - "confidence": 0.81, + "summary": "These PRs are thematically related to multimodal pipeline/processor fixes, but they address distinct code paths: stop-sequence handling, generation kwarg forwarding, model-specific post-processing, input key normalization, audio format validation, and Gemma4 placeholder validation. I did not find any true duplicate pairs.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47051 is the broadest pipeline-level representative: it changes multimodal generation kwarg routing in both AnyToAny and ImageTextToText, matching the shared pipeline-handling theme better than the model-specific fixes.", + "canonical_pr_reason": "PR 47051 is the broadest representative of the cluster: it touches the shared multimodal pipeline argument-routing path and spans both AnyToAny and ImageTextToText, making it the most central PR to anchor this group.", "best_issue_reason": null, - "best_pr_reason": "47051 is the strongest global representative because it touches the core multimodal generation pipeline behavior and spans more than one pipeline implementation.", + "best_pr_reason": "PR 47051 is the best single representative because it targets the common multimodal pipeline layer rather than a single model-specific processor, and its scope is the most central among the set.", "soft_edge_verdicts": [ { "left": "pull_request:47032", - "right": "pull_request:47051", + "right": "pull_request:47045", "accept": false, - "reason": "Same area of code, but different bugs: 47032 fixes stop_sequence -> stop_strings handling, while 47051 routes direct generation kwargs. They are not the same concrete change." + "reason": "Different bugs in different components: image-text-to-text stop-sequence handling vs Voxtral audio-array format handling." }, { "left": "pull_request:47032", - "right": "pull_request:47045", + "right": "pull_request:47051", "accept": false, - "reason": "Unrelated features in different processors: stop-sequence handling for image-text-to-text vs audio-format requirements in Voxtral." + "reason": "Related only at a high level; 47032 fixes explicit stop_sequence conversion, while 47051 is about routing direct generation kwargs through multimodal pipelines." }, { "left": "pull_request:47045", "right": "pull_request:47076", "accept": false, - "reason": "Both involve audio, but one validates Voxtral transcription audio formats and the other validates Gemma4 audio placeholders. Different code paths and fixes." + "reason": "Both are processor changes, but they fix unrelated validation paths: Voxtral audio encoding vs Gemma4 audio/text placeholder construction." }, { "left": "pull_request:47045", "right": "pull_request:47074", "accept": false, - "reason": "Voxtral audio-array encoding and AnyToAny input-key normalization are separate bugs with no shared concrete fix." + "reason": "Distinct multimodal input problems in different code paths; Voxtral audio format encoding is unrelated to AnyToAny key normalization." }, { "left": "pull_request:47051", "right": "pull_request:47074", "accept": false, - "reason": "Both touch AnyToAnyPipeline, but 47051 forwards generation kwargs while 47074 normalizes multimodal input keys. They would not merge as one PR." + "reason": "Shared pipeline file but different underlying changes: kwarg forwarding/generation config handling vs input key normalization for image/video aliases." }, { "left": "pull_request:47051", "right": "pull_request:47060", "accept": false, - "reason": "Different subsystems and bugs: generation-kwarg forwarding in multimodal pipelines vs Qwen Omni batched text post-processing." + "reason": "One is generic pipeline generation-argument plumbing; the other is Qwen Omni batched post-processing decode behavior." }, { "left": "pull_request:47060", "right": "pull_request:47074", "accept": false, - "reason": "Different concrete problems: Qwen Omni output decoding for batched text vs AnyToAny input-key normalization." + "reason": "Different models and different failure modes: Qwen Omni text post-processing vs AnyToAny multimodal input normalization." }, { "left": "pull_request:47060", "right": "pull_request:47076", "accept": false, - "reason": "Qwen Omni batched post-processing and Gemma4 audio placeholder validation are unrelated processor fixes." + "reason": "Separate model-specific processor bugs with no shared concrete code path." }, { "left": "pull_request:47074", "right": "pull_request:47076", "accept": false, - "reason": "AnyToAny input normalization and Gemma4 processor placeholder validation are distinct changes in different code paths." + "reason": "Both are multimodal preprocessing fixes, but they address different inputs and validations in unrelated processors." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index c041ba975812b1fed973633333c5ecc0f5b285c2..ed5c9ae17ebf4543cd45426984cff3a22f404b90 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260705t000031z", + "analysis_id": "hybrid-model-20260705t060758z", "archived_artifacts": { - "hybrid": "snapshots/20260705T000031Z/analysis-runs/hybrid-model-20260705t000031z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260705T000031Z/analysis-runs/hybrid-model-20260705t000031z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-07-05T00:05:54Z", + "published_at": "2026-07-05T06:18:11Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260705T000031Z", + "snapshot_id": "20260705T060758Z", "variant": "hybrid" } diff --git a/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..e8e2e4fbcd9e644e7114fb81f1a098a0eb8ea0d2 --- /dev/null +++ b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json @@ -0,0 +1,3419 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260705T060758Z", + "generated_at": "2026-07-05T06:18:09Z", + "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-46907-7", + "summary": "Cluster of 7 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46907, + "issue_numbers": [], + "pr_numbers": [ + 46907, + 46927, + 46949, + 46951, + 46971, + 47055, + 47056 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46907, + "right_pr_number": 46927, + "code_similarity": 0.209, + "size_similarity": 0.929, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.158, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46907, + "right_pr_number": 46949, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46907, + "right_pr_number": 46951, + "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": 46907, + "right_pr_number": 46971, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46907, + "right_pr_number": 47055, + "code_similarity": 0.925, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 0.5, + "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, + 472 + ] + ] + }, + { + "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "left_ranges": [ + [ + 210, + 218 + ] + ], + "right_ranges": [ + [ + 210, + 218 + ] + ] + } + ] + }, + { + "left_pr_number": 46907, + "right_pr_number": 47056, + "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": 46927, + "right_pr_number": 46949, + "code_similarity": 0.157, + "size_similarity": 0.786, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 46951, + "code_similarity": 0.079, + "size_similarity": 0.393, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 46971, + "code_similarity": 0.162, + "size_similarity": 0.786, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 47055, + "code_similarity": 0.209, + "size_similarity": 0.929, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.158, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46927, + "right_pr_number": 47056, + "code_similarity": 0.079, + "size_similarity": 0.393, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 46951, + "code_similarity": 0.1, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 46971, + "code_similarity": 0.206, + "size_similarity": 1.0, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 47055, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46949, + "right_pr_number": 47056, + "code_similarity": 0.54, + "size_similarity": 0.5, + "file_overlap": 0.5, + "area_overlap": 0.4, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/whisper/modeling_whisper.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/whisper/modeling_whisper.py", + "left_ranges": [ + [ + 309, + 315 + ] + ], + "right_ranges": [ + [ + 306, + 312 + ] + ] + } + ] + }, + { + "left_pr_number": 46951, + "right_pr_number": 46971, + "code_similarity": 0.106, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46951, + "right_pr_number": 47055, + "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": 46951, + "right_pr_number": 47056, + "code_similarity": 0.2, + "size_similarity": 1.0, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46971, + "right_pr_number": 47055, + "code_similarity": 0.169, + "size_similarity": 0.846, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46971, + "right_pr_number": 47056, + "code_similarity": 0.106, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47055, + "right_pr_number": 47056, + "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": [] + } + ] + }, + { + "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-43451-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 43451, + "issue_numbers": [], + "pr_numbers": [ + 43451, + 46836 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43451, + "right_pr_number": 46836, + "code_similarity": 0.261, + "size_similarity": 0.792, + "file_overlap": 0.176, + "area_overlap": 0.0, + "patch_similarity": 0.332, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "src/transformers/conversion_mapping.py", + "src/transformers/models/__init__.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py", + "utils/check_repo.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 1367, + 1374 + ] + ], + "right_ranges": [ + [ + 887, + 894 + ] + ] + }, + { + "filename": "src/transformers/conversion_mapping.py", + "left_ranges": [ + [ + 664, + 696 + ] + ], + "right_ranges": [ + [ + 1420, + 1486 + ] + ] + }, + { + "filename": "src/transformers/models/__init__.py", + "left_ranges": [ + [ + 289, + 295 + ] + ], + "right_ranges": [ + [ + 452, + 458 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 386, + 393 + ], + [ + 816, + 822 + ], + [ + 936, + 942 + ], + [ + 1057, + 1063 + ], + [ + 1189, + 1195 + ] + ], + "right_ranges": [ + [ + 628, + 638 + ], + [ + 905, + 914 + ], + [ + 1112, + 1118 + ], + [ + 1238, + 1244 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 326, + 333 + ], + [ + 1087, + 1093 + ], + [ + 1906, + 1912 + ] + ], + "right_ranges": [ + [ + 501, + 507 + ], + [ + 1107, + 1113 + ] + ] + }, + { + "filename": "utils/check_repo.py", + "left_ranges": [ + [ + 250, + 258 + ], + [ + 333, + 340 + ] + ], + "right_ranges": [ + [ + 267, + 274 + ], + [ + 489, + 496 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46908-5", + "summary": "Cluster of 5 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46926, + "issue_numbers": [], + "pr_numbers": [ + 46908, + 46926, + 47054, + 47057, + 47058 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46908, + "right_pr_number": 46926, + "code_similarity": 0.31, + "size_similarity": 0.8, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 1.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46908, + "right_pr_number": 47054, + "code_similarity": 0.31, + "size_similarity": 0.8, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 1.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46908, + "right_pr_number": 47057, + "code_similarity": 0.565, + "size_similarity": 0.231, + "file_overlap": 1.0, + "area_overlap": 0.565, + "patch_similarity": 0.143, + "shared_filenames": [ + "tests/models/deepseek_v32/test_modeling_deepseek_v32.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/deepseek_v32/test_modeling_deepseek_v32.py", + "left_ranges": [ + [ + 13, + 19 + ], + [ + 251, + 257 + ], + [ + 270, + 276 + ], + [ + 291, + 297 + ], + [ + 332, + 338 + ] + ], + "right_ranges": [ + [ + 13, + 19 + ], + [ + 248, + 260 + ], + [ + 268, + 281 + ], + [ + 290, + 303 + ], + [ + 332, + 345 + ] + ] + } + ] + }, + { + "left_pr_number": 46908, + "right_pr_number": 47058, + "code_similarity": 0.15, + "size_similarity": 0.75, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46926, + "right_pr_number": 47054, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "tests/models/minimax_m3_vl/test_modeling_minimax_m3_vl.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/minimax_m3_vl/test_modeling_minimax_m3_vl.py", + "left_ranges": [ + [ + 14, + 20 + ], + [ + 487, + 493 + ] + ], + "right_ranges": [ + [ + 14, + 20 + ], + [ + 487, + 493 + ] + ] + } + ] + }, + { + "left_pr_number": 46926, + "right_pr_number": 47057, + "code_similarity": 0.058, + "size_similarity": 0.185, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.143, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46926, + "right_pr_number": 47058, + "code_similarity": 0.12, + "size_similarity": 0.6, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47054, + "right_pr_number": 47057, + "code_similarity": 0.058, + "size_similarity": 0.185, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.143, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47054, + "right_pr_number": 47058, + "code_similarity": 0.12, + "size_similarity": 0.6, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47057, + "right_pr_number": 47058, + "code_similarity": 0.083, + "size_similarity": 0.308, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.143, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46894-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46894, + "canonical_pr_number": 46895, + "issue_numbers": [ + 46894, + 46897 + ], + "pr_numbers": [ + 46895, + 46958 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "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 + ] + ] + } + ] + } + ] + }, + { + "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": 0.55, + "canonical_issue_number": 45561, + "canonical_pr_number": 45645, + "issue_numbers": [ + 45561 + ], + "pr_numbers": [ + 45639, + 45645 + ], + "evidence_types": [ + "closing_reference" + ], + "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-47013-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #47013.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 47013, + "canonical_pr_number": 47026, + "issue_numbers": [ + 47013 + ], + "pr_numbers": [ + 47026, + 47046 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47026, + "right_pr_number": 47046, + "code_similarity": 0.308, + "size_similarity": 0.264, + "file_overlap": 0.5, + "area_overlap": 0.229, + "patch_similarity": 0.169, + "shared_filenames": [ + "src/transformers/models/xlstm/modeling_xlstm.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/xlstm/modeling_xlstm.py", + "left_ranges": [ + [ + 816, + 840 + ], + [ + 843, + 851 + ] + ], + "right_ranges": [ + [ + 25, + 39 + ], + [ + 789, + 795 + ], + [ + 808, + 817 + ], + [ + 823, + 846 + ], + [ + 849, + 857 + ], + [ + 1388, + 1394 + ], + [ + 1399, + 1405 + ], + [ + 1461, + 1473 + ], + [ + 1504, + 1522 + ] + ] + } + ] + } + ] + }, + { + "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-47048-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #47065.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 47065, + "canonical_pr_number": 47066, + "issue_numbers": [ + 47065 + ], + "pr_numbers": [ + 47048, + 47066 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47048, + "right_pr_number": 47066, + "code_similarity": 0.3, + "size_similarity": 0.95, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.734, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46970-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46986, + "issue_numbers": [], + "pr_numbers": [ + 46970, + 46986 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46970, + "right_pr_number": 46986, + "code_similarity": 0.138, + "size_similarity": 0.285, + "file_overlap": 0.194, + "area_overlap": 0.038, + "patch_similarity": 0.061, + "shared_filenames": [ + "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", + "src/transformers/models/granitemoehybrid/modeling_granitemoehybrid.py", + "src/transformers/models/lfm2_moe/modeling_lfm2_moe.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/qwen3_5/modeling_qwen3_5.py", + "src/transformers/models/qwen3_5_moe/modeling_qwen3_5_moe.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", + "left_ranges": [ + [ + 140, + 147 + ] + ], + "right_ranges": [ + [ + 30, + 36 + ], + [ + 438, + 444 + ], + [ + 509, + 515 + ], + [ + 534, + 539 + ] + ] + }, + { + "filename": "src/transformers/models/granitemoehybrid/modeling_granitemoehybrid.py", + "left_ranges": [ + [ + 22, + 40 + ], + [ + 364, + 371 + ], + [ + 861, + 943 + ], + [ + 1063, + 1069 + ], + [ + 1074, + 1083 + ] + ], + "right_ranges": [ + [ + 35, + 41 + ], + [ + 731, + 737 + ] + ] + }, + { + "filename": "src/transformers/models/lfm2_moe/modeling_lfm2_moe.py", + "left_ranges": [ + [ + 405, + 412 + ] + ], + "right_ranges": [ + [ + 34, + 40 + ], + [ + 491, + 497 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 222, + 229 + ] + ], + "right_ranges": [ + [ + 38, + 44 + ], + [ + 586, + 592 + ] + ] + }, + { + "filename": "src/transformers/models/qwen3_5/modeling_qwen3_5.py", + "left_ranges": [ + [ + 412, + 417 + ], + [ + 429, + 436 + ], + [ + 825, + 831 + ], + [ + 1220, + 1225 + ] + ], + "right_ranges": [ + [ + 33, + 39 + ], + [ + 437, + 443 + ] + ] + }, + { + "filename": "src/transformers/models/qwen3_5_moe/modeling_qwen3_5_moe.py", + "left_ranges": [ + [ + 33, + 39 + ], + [ + 409, + 414 + ], + [ + 426, + 433 + ], + [ + 862, + 875 + ], + [ + 909, + 915 + ], + [ + 1292, + 1320 + ], + [ + 1328, + 1333 + ] + ], + "right_ranges": [ + [ + 33, + 39 + ], + [ + 434, + 440 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [ + { + "cluster_id": "cluster-46894-4", + "canonical_issue_number": 46894, + "duplicate_issue_numbers": [ + 46897 + ], + "reason": "Issues in cluster-46894-4 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." + } + ], + "duplicate_prs": [ + { + "cluster_id": "cluster-41211-3", + "canonical_pr_number": 44339, + "duplicate_pr_numbers": [ + 41356 + ], + "target_issue_number": 41211, + "reason": "PRs in cluster-41211-3 are treated as duplicates because they converge on issue #41211 with closing_reference evidence." + }, + { + "cluster_id": "cluster-43240-3", + "canonical_pr_number": 43251, + "duplicate_pr_numbers": [ + 43254 + ], + "target_issue_number": 43240, + "reason": "PRs in cluster-43240-3 are treated as duplicates because they converge on issue #43240 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43451-2", + "canonical_pr_number": 43451, + "duplicate_pr_numbers": [ + 46836 + ], + "target_issue_number": null, + "reason": "PRs in cluster-43451-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-43698-3", + "canonical_pr_number": 43779, + "duplicate_pr_numbers": [ + 43816 + ], + "target_issue_number": 43698, + "reason": "PRs in cluster-43698-3 are treated as duplicates because they converge on issue #43698 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43979-11", + "canonical_pr_number": 44007, + "duplicate_pr_numbers": [ + 43996, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "target_issue_number": 43979, + "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45561-3", + "canonical_pr_number": 45645, + "duplicate_pr_numbers": [ + 45639 + ], + "target_issue_number": 45561, + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference 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-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-46894-4", + "canonical_pr_number": 46895, + "duplicate_pr_numbers": [ + 46958 + ], + "target_issue_number": 46894, + "reason": "PRs in cluster-46894-4 are treated as duplicates because they converge on issue #46894 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46907-7", + "canonical_pr_number": 46907, + "duplicate_pr_numbers": [ + 46927, + 46949, + 46951, + 46971, + 47055, + 47056 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46907-7 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46908-5", + "canonical_pr_number": 46926, + "duplicate_pr_numbers": [ + 46908, + 47054, + 47057, + 47058 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46908-5 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46970-2", + "canonical_pr_number": 46986, + "duplicate_pr_numbers": [ + 46970 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46970-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47013-3", + "canonical_pr_number": 47026, + "duplicate_pr_numbers": [ + 47046 + ], + "target_issue_number": 47013, + "reason": "PRs in cluster-47013-3 are treated as duplicates because they converge on issue #47013 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47048-3", + "canonical_pr_number": 47066, + "duplicate_pr_numbers": [ + 47048 + ], + "target_issue_number": 47065, + "reason": "PRs in cluster-47048-3 are treated as duplicates because they converge on issue #47065 with closing_reference, soft_similarity evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-46894-4", + "issue_number": 46894, + "reason": "Issue #46894 is the strongest global issue candidate because it is open, belongs to a cluster with 4 artifacts, and carries 0 discussion comments plus 6 inbound references.", + "score": 146.0 + }, + "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/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..05cdcd9dd4891cf70c58ca12d0fad739d169f6ac --- /dev/null +++ b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,26438 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260705T060758Z", + "generated_at": "2026-07-05T06:18:09Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43425", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44805", + "issue:44841", + "issue:44918", + "issue:45081", + "issue:45161", + "issue:45237", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45800", + "issue:45823", + "issue:45859", + "issue:45874", + "issue:45901", + "issue:45902", + "issue:46015", + "issue:46165", + "issue:46294", + "issue:46468", + "issue:46515", + "issue:46661", + "issue:46869" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:46468", + "issue:45237|issue:45507", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:35545|issue:36010", + "issue:44509|issue:44512", + "issue:44918|issue:46165", + "issue:46165|issue:46869", + "issue:44805|issue:45823", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:46294", + "issue:44805|issue:45161", + "issue:46165|issue:46468", + "issue:44560|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43039|issue:45161", + "issue:43975|issue:44457", + "issue:44292|issue:44457", + "issue:44841|issue:46661", + "issue:43716|issue:46869", + "issue:24643|issue:30277", + "issue:34928|issue:44805", + "issue:45685|issue:46869", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45874|issue:46294", + "issue:44292|issue:45237", + "issue:39290|issue:45081", + "issue:45823|issue:45874", + "issue:45800|issue:46515", + "issue:34928|issue:41720", + "issue:43425|issue:44292", + "issue:43828|issue:45161", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:46468", + "issue:44292|issue:44918", + "issue:45507|issue:45685", + "issue:43716|issue:45685", + "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": "82f9a1ec9fbd68097d9f773b7bb2353a92764a69633b568909e17e935b80ef31", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 46572, + "estimated_input_tokens": 11643, + "estimated_eval_tokens": 23542 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous, but there are two conservative duplicate subgroups: a ZeRO-3 index-out-of-bounds/loading-failure subgroup and a docs-cleanup subgroup about v5 removals. The ZeRO-3 items look like near-duplicates, not provably exact duplicates from the packet alone.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 43638 is the clearest representative of the ZeRO-3 'index 0 is out of bounds' subgroup. It is a likely near-duplicate of 43596 and closely related to 41762, but the packet only supports a conservative same-bug grouping, not an exact-duplicate claim.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43638 is the best global issue representative because it names the core ZeRO-3 index-out-of-bounds failure most specifically and sits at the center of the strongest near-duplicate cluster in the packet.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different bugs: Qwen3.5 config attribute error vs streaming chat/completions failure." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different model/backend problems; no shared concrete fix path." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Both are runtime failures, but the model and failure mechanism differ." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Autocast dtype mismatch is not the same as the AMD GPU device-map failure." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Different models and different failure modes; not one bug." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Invalid method call in GraniteMoEHybrid and AMD GPU failure are unrelated." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collectives mismatch and ONNX export error are unrelated." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export and Qwen3Moe FP8 failure are different code paths." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Export-to-ONNX vs auto device-mapping CUDA assert are different bugs." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Task crash and docs cleanup are not the same underlying issue." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ONNX export issue and GenerationMixin import error are distinct." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": true, + "reason": "Both are docs cleanup for v5 removals, so they can be fixed together." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "TRL embedding unpacking bug and resume-training/trust-remote-code bug are unrelated." + }, + { + "left": "issue:46165", + "right": "issue:46869", + "accept": false, + "reason": "Different training/device-map problems with no shared concrete fix." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Mask/index shape mismatch and Gemma4 device mismatch are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both are index-ish errors, but the ZeRO-3 loading bug is a different code path from the mask-shape mismatch." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 loading failure and resume-from-local-model behavior are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Same broad error family, but not the same concrete bug or code path." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal input issues: mask indexing vs video token/feature mismatch." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Generic index error and GPT-OSS tensor-parallel issue are unrelated." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Resume-training bug and broken example documentation are unrelated." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Video StopIteration bug and training resume bug are different." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "Processor preset bug and ONNX export error are unrelated." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Both involve DeepSpeed, but the concrete failures are different." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 index-out-of-bounds loading failure; 43596 is a likely near-duplicate of 43638." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross-entropy call and GPT-OSS tensor-parallel bug are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Incorrect detokenization and LoRA save/load output mismatch are different issues." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen NVFP4 runtime error and LoRA merge/save mismatch are unrelated." + }, + { + "left": "issue:44841", + "right": "issue:46661", + "accept": false, + "reason": "Processor failure and remote-code config_class error are not the same bug." + }, + { + "left": "issue:43716", + "right": "issue:46869", + "accept": false, + "reason": "Different models and backend/device-map problems." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "DeepSpeed training errors differ: weight-shape assertion vs collective mismatch." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpoint recompute mismatch is not the same as a mask/indexing error." + }, + { + "left": "issue:45685", + "right": "issue:46869", + "accept": false, + "reason": "MPS histogram backend issue and multi-GPU device-map issue are unrelated." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": false, + "reason": "Same model family, but the CPU offload incompatibility and uninitialized-memory materialization are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 Bert loading bug and GPT-OSS tensor-parallel issue are unrelated." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Generic torch compatibility issue vs specific Qwen NVFP4 runtime error." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch and MPS integer-histogram limitation are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "Grouped_mm tracing failure and image-preprocessor dtype mismatch are different bugs." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Different model/runtime failures with no shared fix." + }, + { + "left": "issue:43716", + "right": "issue:46468", + "accept": false, + "reason": "Different models and failure modes." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Different Qwen variants and different bugs." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Invalid method call in GraniteMoEHybrid and unsupported MPS op are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Different models and backend problems." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRL embedding unpacking bug and GPT-OSS tensor-parallel issue are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger Kernel cross-entropy issue and TRL SFT trainer bug are unrelated." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 'index 0 out of bounds' loading failure, just reported on a different model." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": true, + "reason": "Both are docs references left behind after v5 removals." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": true, + "reason": "Both describe the same torch 2.4.1 incompatibility." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Both are Gemma4 bugs, but the device mismatch and reverse-embedding lookup are different problems." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Different models and failures: Gemma3 config attribute vs Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Repeated pair is still unrelated: autocast mismatch vs unsupported MPS op." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger Kernel cross_entropy issue and mask/index shape error are not the same bug." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall clustering is plausible, but the writeup overstates the evidence in a few places. In particular, calling issue 43596 an \"exact match\" to 43638 and describing an \"exact duplicate relationship\" is stronger than what the packet supports: the titles are highly similar and the issues look like likely duplicates, but there is no explicit duplicate linkage or body text here proving exact sameness. Please soften that language to \"near-duplicate\" / \"likely duplicate\" and keep the subgroup claim conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 2, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43866|issue:46402", + "issue:43296|issue:46304", + "issue:41762|issue:43596", + "issue:39290|issue:41720", + "issue:43901|issue:44512", + "issue:44589|issue:46304", + "issue:45237|issue:45901", + "issue:43039|issue:44560", + "issue:43975|issue:46015", + "issue:45823|issue:46294", + "issue:44841|issue:45823", + "issue:43866|issue:46772", + "issue:44863|issue:46015", + "issue:45020|issue:46304", + "issue:45812|issue:45920", + "issue:43866|issue:44863", + "issue:44292|issue:46015", + "issue:43872|issue:46304", + "issue:39290|issue:43531", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43854|issue:43866", + "issue:43425|issue:45070", + "issue:43975|issue:45901", + "issue:43792|issue:45070", + "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:43404|issue:43866", + "issue:45507|issue:46869", + "issue:41093|issue:41762", + "issue:45399|issue:46661", + "issue:43572|issue:45070", + "issue:45800|issue:45823" + ], + "prepared_review_unit_hash": "68c40c7a65f2cf3fcaf76bdc02d931deb39307bbd2bbd871b1dc2910f580424b", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 40042, + "estimated_input_tokens": 10011, + "estimated_eval_tokens": 20278 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous issue noise. I found two likely duplicate clusters: the DeepSpeed ZeRO-3 load/indexing failures and the v5 AutoTokenizer wrong-token-id regression. All other soft pairs look unrelated.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45920 is the broadest and clearest representative of the tokenizer regression cluster; it covers multiple model families and is more general than issue:45812.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best global issue candidate here because it most clearly captures a recurring, cross-model v5 tokenizer bug that the paired report also points to.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Both mention torch/version breakage, but they describe different incompatibilities and different failure modes." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "Different models and different bugs: corrupted checkpoint vs missing text-encoder weights." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Different stack and symptom; vLLM/PaddleOCR-VL load failure is unrelated to the USE_DISTRIBUTED=0 PyTorch support break." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Both report the same ZeRO-3 initialization/load path producing an index-out-of-bounds on empty tensors during model loading." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Gemma3 sliding_window_pattern missing in vLLM and Qwen3 auto-device-map cuda assert are different code paths and failures." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs-related, but they cover different outdated references and are not the same documentation defect." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Unrelated runtime errors: missing Float8 storage vs PyTorch build/distributed support break." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Different models/features and different failures; no evidence of a shared underlying bug." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger Kernel cross_entropy dispatch and Qwen3-vl video StopIteration are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "A detokenization regression is unrelated to outdated RAG examples." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Device-map forward mismatch and chat-template video token mismatch are different issues." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Processor failure for Voxtral and Gemma4 device mismatch are distinct bugs." + }, + { + "left": "issue:43866", + "right": "issue:46772", + "accept": false, + "reason": "Corrupted checkpoint and NVFP4 compatibility are different problems." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "NemotronH checkpoint loading failure is unrelated to outdated RAG examples." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Both are regressions in recent versions, but they affect different subsystems and error modes." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Both describe the same v5 tokenizer regression causing wrong token IDs across multiple model families; 45920 is the broader report." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Both are model-loading failures, but they concern different models and likely different root causes." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime failure is unrelated to RAG documentation drift." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes Int8Params API break is unrelated to the PyTorch distributed build issue." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but the models, stack, and failures are different enough that this does not look like one bug." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Different error shapes and different contexts; one is a mask-length mismatch, the other a zero-3/empty tensor index error." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper runtime failure and table-question-answering crash are unrelated." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Different model-specific loading problems with no clear shared root cause." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch 2.10 compatibility and pydantic PretrainedConfig breakage are separate compatibility issues." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization regression and table QA crash are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper load/runtime failure does not match the pydantic config regression." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Both involve Qwen3.5 chat behavior, but one is transformers chat and the other is streaming chat/completions; too different to merge." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "ZeRO-3 load/index error is unrelated to the torch/transformers version incompatibility." + }, + { + "left": "issue:43541", + "right": "issue:46869", + "accept": false, + "reason": "Mixtral torch-dynamo grouped_mm tracing failure is unrelated to YOLO multi-GPU device-map auto failure." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Qwen NVFP4 runtime issue and table QA crash are not the same bug." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Different models and different failure modes: missing config attr in vLLM versus Gemma4 forward device mismatch." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Qwen3.5 streaming API failure is unrelated to the MPS histogram_int error." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin and a Gemma3 sliding_window config issue are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Mixtral grouped_mm tracing and MPS histogram_int errors are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Different MoE-related symptoms, but different models, execution modes, and likely root causes." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Processor loading failure and template-node compilation error are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 is unrelated to a corrupted Ovis2 checkpoint." + }, + { + "left": "issue:45507", + "right": "issue:46869", + "accept": false, + "reason": "GraniteMoEHybrid invalid-method calls and YOLO multi-GPU device-map failure are different bugs." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Mask-shape mismatch and ZeRO-3 zero-sized index error are different failure signatures." + }, + { + "left": "issue:45399", + "right": "issue:46661", + "accept": false, + "reason": "Flash-attn fallback gating and remote-code config_class AttributeError are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is a different config issue than the pydantic PretrainedConfig regression." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch/transformers compatibility break is unrelated to Gemma4 device placement mismatch." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative. The two accepted duplicate clusters are supported by the titles: ZeRO-3 load/index-out-of-bounds (41762/43596) and the v5 AutoTokenizer wrong-token-id regression (45812/45920). The remaining soft-edge rejects are generally defensible as different models, stacks, or failure modes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 3, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:39186", + "issue:40990", + "issue:41093", + "issue:42915", + "issue:43054", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43873", + "issue:43994", + "issue:44291", + "issue:44360", + "issue:44387", + "issue:44479", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44805", + "issue:44863", + "issue:44918", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45161", + "issue:45183", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45632", + "issue:45698", + "issue:45759", + "issue:45800", + "issue:45902", + "issue:46165", + "issue:46294", + "issue:46463", + "issue:46468", + "issue:46640", + "issue:46661", + "issue:46869" + ], + "soft_pairs": [ + "issue:43296|issue:43366", + "issue:44805|issue:44918", + "issue:43541|issue:45161", + "issue:43828|issue:45464", + "issue:44387|issue:45005", + "issue:43366|issue:45464", + "issue:44291|issue:45070", + "issue:43531|issue:45070", + "issue:41093|issue:43596", + "issue:43716|issue:46294", + "issue:45020|issue:45800", + "issue:43054|issue:43994", + "issue:44479|issue:46294", + "issue:46468|issue:46869", + "issue:45759|issue:45902", + "issue:43596|issue:46165", + "issue:46640|issue:46661", + "issue:43716|issue:44918", + "issue:39186|issue:44918", + "issue:43645|issue:45399", + "issue:40990|issue:44960", + "issue:45632|issue:45698", + "issue:43404|issue:45005", + "issue:42915|issue:45399", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:44530|issue:45005", + "issue:44492|issue:44512", + "issue:44863|issue:45005", + "issue:45183|issue:46463", + "issue:44918|issue:45507" + ], + "prepared_review_unit_hash": "9fbe597630892ba93fb13dd2ca5fca6a3fd4f54febb702235f99fa86dd1ff67d", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36240, + "estimated_input_tokens": 9060, + "estimated_eval_tokens": 18376 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a collection of distinct Transformers bug reports spanning model loading, multimodal preprocessing, quantization, cache behavior, and validation errors. The soft pairs share broad themes but do not look like the same underlying defect, so none are accepted.", + "confidence": 0.9, + "canonical_issue_reason": "issue:45020 is the broadest, clearest regression report in the set: recent versions breaking remote_code models. It is concrete, user-facing, and representative of the cluster\u2019s recurring loading regressions.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the most generally useful canonical issue because it captures a high-impact, cross-model failure mode with clear actionable scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "Different bugs: PaddleOCR-VL model load failure versus gpt-oss GGUF architecture support." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Both are shape-related failures, but one is a mask/tensor index mismatch and the other is a TRL SFT unpacking issue." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Different code paths: torch dynamo/grouped_mm tracing failure versus tensor-parallel-only GPT-OSS MoE behavior." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Unrelated symptoms: autocast dtype mismatch in Phi-tiny-MoE versus streaming chat/completions failure in Qwen3.5." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "One is an int4 quantization memory regression; the other is a tied-weights issue on translation models." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GGUF architecture support is unrelated to Qwen3.5 streaming inference in chat/completions." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Different regressions: unexpected _is_hf_initialized arg in init_empty_weights versus pydantic PretrainedConfig field breakage." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is not the same bug as the v5.4.0 pydantic model break." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both are IndexErrors, but the failing shapes and initialization contexts are different." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Both involve multimodal models, but one is an image-preprocessor dtype mismatch and the other is a video feature/token count mismatch." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote-code compatibility breakage is not the same as the torch 2.4.1 / transformers 5.8.0 incompatibility report." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but one reports degraded text embeddings and the other nonsensical AutoModel/pipeline outputs; not clearly the same defect." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-related, but the regression scope and failure modes differ." + }, + { + "left": "issue:46468", + "right": "issue:46869", + "accept": false, + "reason": "Official Nemotron example breakage and YOLO multi-GPU device-map failure are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Both touch Qwen multimodal internals, but one is text_config unwrapping and the other is uninitialized inv_freq after to_empty materialization." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "Zero3/BERT indexing failure is unrelated to resuming training with locally downloaded remote_code models." + }, + { + "left": "issue:46640", + "right": "issue:46661", + "accept": false, + "reason": "Flash Attention 2 fallback issues and remote-code model_class/config_class errors are different code paths." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Different failures: multimodal dtype mismatch versus TRL SFT embedding unpacking." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP weight dimensionality error is unrelated to the TRL SFT trainer embedding issue." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Jupyter custom-model initialization and FlashAttention fallback blocking are distinct regressions." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-specific issues." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both involve remote_code/custom-module loading, but one is a leaf-directory cache collision while the other is wrong module loading after save_pretrained; not enough to merge." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Mistral3 lm_head tie-breaking affects a different model and path than the translation-model tied-weights regression." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "FineGrainedFP8Config failure and FlashAttention fallback blocking are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is a different issue from tied weights on translation models." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "A DSA indexer activation bug and a docs typo are not the same underlying change." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash on Qwen3.5 is unrelated to tied-weight handling in translation models." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A cache-strategy typo and a removed run-command doc reference are separate documentation issues." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading failure is unrelated to tied weights in translation models." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": false, + "reason": "The generic input_ids/inputs_embeds XOR validation issue and T5 decoder-input validation are related in spirit but not the same bug." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "Qwen3.5 TRL embedding unpacking and GraniteMoEHybrid invalid method calls are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 4, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:24643", + "issue:28530", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:41553", + "issue:41762", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43493", + "issue:43643", + "issue:43644", + "issue:43782", + "issue:43824", + "issue:43828", + "issue:43856", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44292", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44910", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45399", + "issue:45520", + "issue:45663", + "issue:45701", + "issue:45859", + "issue:45902", + "issue:45920", + "issue:45987", + "issue:46489", + "issue:46491", + "issue:46858", + "issue:46899", + "issue:46932" + ], + "soft_pairs": [ + "issue:43975|issue:44292", + "issue:46489|issue:46491", + "issue: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:45920|issue:46489", + "issue:43054|issue:45701", + "issue:38175|issue:43054", + "issue:45987|issue:46858", + "issue:43950|issue:45902", + "issue:44910|issue:45859", + "issue:43054|issue:43493", + "issue:42947|issue:44387", + "issue:45663|issue:45902", + "issue:41553|issue:43883", + "issue:44403|issue:44451", + "issue:43975|issue:45520", + "issue:24643|issue:28530", + "issue:44403|issue:44488", + "issue:43643|issue:45020", + "issue:42947|issue:43856", + "issue:41762|issue:42915", + "issue:44991|issue:45399", + "issue:41553|issue:43881", + "issue:43824|issue:43872", + "issue:43782|issue:44960" + ], + "prepared_review_unit_hash": "6512e6351e1c37c8217add90b99bb2362ffa3fd98d60f5214ed122826cd52fe4", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36287, + "estimated_input_tokens": 9072, + "estimated_eval_tokens": 18400 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is still mostly a mix of unrelated bug reports, but the non-persistent-buffer regression is a small three-way duplicate cluster: #43644, #43950, and #44534 all describe the same Transformers v5 corruption of `register_buffer(persistent=False)` state. The other soft-edge candidates remain too broad or too code-path-specific to merge.", + "confidence": 0.89, + "canonical_issue_reason": "Issue #43950 is the best canonical issue because it states the regression at the right level of specificity (`from_pretrained()` corrupts non-persistent buffers in Transformers 5.x) while cleanly subsuming the near-duplicate reports #43644 and #44534.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43950 is the strongest representative of the duplicate buffer-corruption cluster: it is the most explicit about the `from_pretrained()` failure mode and general enough to cover the paraphrased reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Both are loading/tokenization failures, but they involve different models and likely different code paths; not the same bug." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different failure modes; too different to merge." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Same model family only in a broad sense; one is a docs/example mismatch with quantization failure, the other is a text-embedding quality regression." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "These are the same non-persistent-buffer regression in Transformers v5, described with slightly different wording." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "A trainer API compatibility error and a DeepSpeed ZeRO-3 loading index error are unrelated failures." + }, + { + "left": "issue:43643", + "right": "issue:43950", + "accept": false, + "reason": "One is missing fields from `AutoConfig.from_pretrained(..., trust_remote_code=True)`; the other is buffer corruption during `from_pretrained()`. Different problems." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "One is a bitsandbytes parameter signature incompatibility; the other is a missing storage class error. Not the same bug." + }, + { + "left": "issue:46899", + "right": "issue:46932", + "accept": false, + "reason": "These are closely related Gemma4 quantization reports, but #46932 reads like a follow-up/fix note rather than a clearly separate duplicate of the exact same defect." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Different symptoms and subsystems: dtype mismatch with autocast vs a flash-attn crash from 3D `position_ids` handling." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Both involve tokenizer regressions, but they affect different model families and different failure modes." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "Both are about tokenization/behavior changes, but one is a SigLIP2 embedding-quality issue and the other is a broad version-dependent tokenization change." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Related SigLIP/SigLIP2 surface area, but one is zero probabilities and the other is worse text embeddings; not the same concrete bug." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": false, + "reason": "Both concern `get_seq_length()`, but one is an API contract mismatch and the other is backend-specific tensor arithmetic breakage." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "Both involve non-persistent buffers, but #45902 is a specific `inv_freq` uninitialized-memory case after `to_empty(cuda)`, not the same regression as #43950." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "One is a flash-attention crash due to 3D `position_ids`; the other is CPU offload incompatibility in rotary embedding. Different code paths." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Both are SigLIP2-related, but one reports worse embeddings and the other compares HF vs JAX implementation discrepancy; not enough to merge." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Both mention memory, but one is gradient checkpointing with PEFT LoRA and the other is int4 quantization reserved-memory growth. Different defects." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Gemma4 FSDP2 shared-state rebuilding and Qwen3.5 non-persistent buffer initialization are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "A bad tokenizer import error for Voxtral and a missing tied-weights attribute for Molmo are different issues." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "One is generic load-time noise; the other is inability to load a specific model. Not the same underlying bug." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Different models and different failure types: detokenization regression vs `import_utils` KeyError on Python 3.13." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "Training DeepSpeed weight-shape runtime error and early-stopping metric misconfiguration are unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but one is noisy logging and the other is a model loading failure for a specific checkpoint." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Both mention remote code / config loading regressions, but one is missing fields in `AutoConfig` return objects while the other is a broader compatibility report." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Both are memory-efficiency concerns, but one is PEFT/gradient checkpointing and the other is Qwen3 MoE training memory usage." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "One is a ZeRO-3 load-time index error for Gemma3; the other is a FineGrainedFP8Config failure for Qwen3 MoE." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer load failure for a specific model and flash-attn fallback logic are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Both are model-loading issues, but one is an AutoTokenizer import error and the other is a `glm-4v-9b` loading failure; no shared root cause is evident." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "An unavailable model class import and a bitsandbytes parameter incompatibility are different failures." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight-only loading error and a generic GLM5 issue are not the same bug." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is mostly conservative, but it overstates the duplicate structure. The packet also contains issue 43644 ('Transformers 5.0.0 fills non-persistent buffers with junk'), which is essentially the same non-persistent-buffer regression family as 43950/44534 and is not accounted for in the canonicalization. Because of that, the claim that there is only one clear duplicate pair and that #43950 is the best/general representative is not fully grounded in the packet. The rest of the soft-edge rejections look reasonably conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 5, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43940", + "issue:43950", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44479", + "issue:44509", + "issue:44560", + "issue:44661", + "issue:44805", + "issue:44863", + "issue:44928", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45305", + "issue:45313", + "issue:45356", + "issue:45399", + "issue:45488", + "issue:45588", + "issue:45685", + "issue:45715", + "issue:45759", + "issue:45823", + "issue:45923", + "issue:46315", + "issue:46489", + "issue:46661", + "issue:46869", + "issue:46894", + "issue:46897" + ], + "soft_pairs": [ + "issue:45685|issue:45823", + "issue:45081|issue:45399", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:42491|issue:44368", + "issue:43881|issue:43883", + "issue:45020|issue:45356", + "issue:45715|issue:46315", + "issue:43872|issue:46869", + "issue:45759|issue:46661", + "issue:43827|issue:44509", + "issue:43872|issue:44291", + "issue:45488|issue:46489", + "issue:43404|issue:43950", + "issue:46894|issue:46897", + "issue:45020|issue:46661", + "issue:43792|issue:44220", + "issue:43257|issue:45305", + "issue:42947|issue:45923", + "issue:36331|issue:41093", + "issue:43940|issue:45313", + "issue:43854|issue:44863", + "issue:41553|issue:44991", + "issue:44928|issue:45588", + "issue:44479|issue:44560", + "issue:41093|issue:44805", + "issue:24643|issue:30064" + ], + "prepared_review_unit_hash": "c6e1effece72fdd07ee8ec6048f30cca98be4ea18dda002bf6a33280980c6fef", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36752, + "estimated_input_tokens": 9188, + "estimated_eval_tokens": 18632 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are false positives across unrelated subsystems, but a few are genuine duplicates: docs cleanup after v5 pipeline removals, the `_is_hf_initialized` init-empty-weights regression, Qwen3 ZeRO-3 weight-loading failures, and the training-loss double-shift bug.", + "confidence": 0.76, + "canonical_issue_reason": "issue:46894 is the clearest umbrella report for the training-loss double-shift regression and is the better representative of that duplicate pair than the narrower Florence2 report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46894 has the broadest, most reusable description of the underlying bug pattern and the strongest case for serving as the cluster\u2019s canonical issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "Different bugs: MPS histogram/int issue vs device_map forward mismatch." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Different code paths: tokenizer regex crash vs flash-attn fallback gating." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Broad remote_code regression vs Gemma4 PLE device placement bug." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Related area, but not the same concrete failure; add-new-model-like tokenizer mapping bug vs generic remote_code breakage." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Different models and symptoms: autocast dtype mismatch vs device placement mismatch." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA/Qwen3_moe compatibility issue vs tied-embeddings warning; not the same bug." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Both model-loading failures, but different underlying problems and model classes." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Generic remote_code breakage vs Kimi tokenizer codec/regex regression." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "Both tokenizer performance regressions, but different APIs and likely different hot paths." + }, + { + "left": "issue:43872", + "right": "issue:46869", + "accept": false, + "reason": "Unrelated: bitsandbytes/_is_hf_initialized incompatibility vs YOLO multi-GPU device map failure." + }, + { + "left": "issue:45759", + "right": "issue:46661", + "accept": false, + "reason": "Different remote-code loader bugs: text_config unwrapping vs missing config_class." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same docs regression: both report removed v5 pipeline tasks (summarization/translation/text2text-generation) still referenced in documentation." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same `_is_hf_initialized` TypeError during model loading/init_empty_weights; 44291 is the generalized report of the same regression." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Same broad tokenizer area, but not enough evidence they are the same underlying bug or fix target." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Different problems: lm_head tying bug vs silent corruption of non-persistent buffers." + }, + { + "left": "issue:46894", + "right": "issue:46897", + "accept": true, + "reason": "Same training-loss double-shift bug pattern; Florence2 looks like a model-specific instance of the same regression." + }, + { + "left": "issue:45020", + "right": "issue:46661", + "accept": false, + "reason": "Not the same concrete bug; broad remote_code regressions should not be merged with the specific config_class AttributeError report." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Both audio/Whisper-related, but the titles point to different failures and code paths." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Different issues: Qwen3 MOE weight conversion/loading vs gradient averaging under DeepSpeed GAS." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "LoRA gradient-checkpointing effectiveness vs missing supports_gradient_checkpointing flag; related theme, different bug." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Both mention shape errors, but the model/training contexts and exact mismatches are different." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": true, + "reason": "Same DeepSpeed ZeRO-3 weight-loading failure pattern across Qwen3 family variants; likely one loader regression." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures for different models; too little evidence of a shared root cause." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer failures: bad Voxtral error message vs est-roberta tokenizer loading breakage." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Unrelated: RLHF gradient explosion/SDPA fallback vs flash_attention s_aux=None crash." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are video regressions, but the affected models and reported failures differ enough to avoid merging." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": false, + "reason": "Same generic error shape, but different reports and likely different underlying causes." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Completely different subsystems: DeepSpeed training error vs image processor segmentation-map failure." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet, and the accepted soft-edge pairs are conservative and supported by the titles: the docs regression pair, the `_is_hf_initialized` init-empty-weights regression, the Qwen3 ZeRO-3 weight-loading pair, and the training-loss double-shift pair. The false verdicts also read appropriately cautious given only 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": 26, + "review_unit_index": 6, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:41553", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43720", + "issue:43824", + "issue:43828", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44488", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44683", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:45030", + "issue:45325", + "issue:45584", + "issue:45663", + "issue:45672", + "issue:45685", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45902", + "issue:46402", + "issue:46489", + "issue:46515", + "issue:46858", + "issue:46899" + ], + "soft_pairs": [ + "issue:44530|issue:44863", + "issue:44863|issue:46402", + "issue:43366|issue:43828", + "issue:41553|issue:44488", + "issue:43366|issue:43531", + "issue:45325|issue:45902", + "issue:43596|issue:46515", + "issue:44155|issue:45584", + "issue:43366|issue:44589", + "issue:45663|issue:45823", + "issue:43638|issue:46515", + "issue:42915|issue:43957", + "issue:43950|issue:45834", + "issue:43824|issue:43957", + "issue:43716|issue:44560", + "issue:43645|issue:43646", + "issue:44910|issue:46858", + "issue:44560|issue:45685", + "issue:42617|issue:43366", + "issue:43550|issue:44910", + "issue:44560|issue:44918", + "issue:36010|issue:42915", + "issue:44683|issue:46515", + "issue:45812|issue:46489", + "issue:45823|issue:46515", + "issue:44661|issue:45823", + "issue:38175|issue:43994", + "issue:44560|issue:44805", + "issue:45672|issue:46899", + "issue:43632|issue:45834", + "issue:44877|issue:45030", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:42491|issue:43720", + "issue:44661|issue:44805" + ], + "prepared_review_unit_hash": "57c57bfc0ad08b2db0944f78527dec842176923827c2dd08b150602b5beecba4", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 38360, + "estimated_input_tokens": 9590, + "estimated_eval_tokens": 19436 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous: one clear duplicate pair around custom model initialization, one likely duplicate pair around SIGLIP2 output corruption, and many unrelated model-specific regressions.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43632 is the broadest representative of the loading/initialization regression subgroup, centered on the Transformers v5 `_is_hf_initialized` breakage; the other nearby reports are either more specific variants or unrelated.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43632 is the best umbrella issue in this set because it captures the main cross-cutting Transformers v5 model-initialization regression theme, even though the overall cluster contains many unrelated bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different failures: Qwen3.5 PagedAttentionCache linear_attention crash vs NemotronH checkpoint loading failure." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Both are loading-related, but one is NemotronH checkpoint loading and the other is Sam3Model text-encoder weights not loading; not the same bug." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "Unrelated: GGUF gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Both mention tokenizer/model loading, but the affected models and symptoms differ; no clear shared code-path." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Unrelated: GGUF architecture support vs Qwen3-MoE sliding_window behavior." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Different Qwen model bugs: temporal position-id scaling vs uninitialized inv_freq after meta materialization." + }, + { + "left": "issue:43596", + "right": "issue:46515", + "accept": false, + "reason": "Deepspeed Zero3/BertModel index error is not the same as a general torch 2.4.1 incompatibility report." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Different audio bugs: track leakage in batched inference vs empty-transcription generation failure." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated: GGUF support for gpt-oss vs missing Float8 storage during loading." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Both are Gemma4-related, but one is FSDP2 shared_kv_states rebuild and the other is device_map forward mismatch; different problems." + }, + { + "left": "issue:43638", + "right": "issue:46515", + "accept": false, + "reason": "Specific Zero3/BertModel index error is not the same as a generic torch version incompatibility." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Different root issues: Qwen3Moe FP8 failure vs meta-device loading regression." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer corruption and Kosmos2.5 long-OCR index error are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "ImportError for Qwen2_5_VLForConditionalGeneration is unrelated to meta-device loading breakage." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different model families and symptoms: Mistral-3 dtype mismatch vs Qwen3-vl video StopIteration." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same underlying bug: Transformers 5.0.0 breaks custom model initialization; the notebook-specific report is just a more specific instance of the general issue." + }, + { + "left": "issue:44910", + "right": "issue:46858", + "accept": false, + "reason": "Both involve Qwen3.5/cache behavior, but one is a Flash Attention position_ids crash and the other is a static-cache slice arithmetic issue; not the same bug." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Unrelated: Qwen3-vl embedding video StopIteration vs MPS histogram support for MoE integers." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py runtime problem is unrelated to GGUF gpt-oss support." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different models and code paths: Bamba SDPA torch.compile failure vs Qwen3.5 Flash Attention memory access crash." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Both are Qwen-related, but one is video embedding StopIteration and the other is TRL SFT unpacking failure; not the same defect." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "ImportError for GenerationMixin and Qwen3Moe FP8 failure are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:46515", + "accept": false, + "reason": "Torch>=2.9 flex_attention compilation failure is the opposite end of the version spectrum from a torch 2.4.1 incompatibility report." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but on different model families and likely different fixes; not the same underlying bug." + }, + { + "left": "issue:45823", + "right": "issue:46515", + "accept": false, + "reason": "Gemma4 PLE device mismatch is unrelated to a generic torch version incompatibility." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Add-new-model-like / tokenizer mapping issue is unrelated to Gemma4 device placement during forward." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Likely the same SIGLIP2 regression: both report incorrect/nonsensical outputs for google/siglip2-base-patch16-224." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different Qwen3-vl/OCR-related errors with distinct symptoms and no shared concrete code-path." + }, + { + "left": "issue:45672", + "right": "issue:46899", + "accept": false, + "reason": "Both are Gemma4 multimodal quantization issues, but one is torch.finfo on uint8 audio modules and the other is unsafe dtype casting in embedders; too different to merge." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "Broad v5 initialization regression and Kosmos2.5 long-OCR index error are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Both mention config loading/validation, but they affect different model configs and do not clearly share the same bug." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Zero3/BertModel index error is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different failures: deepspeed zero3 init index error vs new-model helper rejection in tokenizer mappings." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3_moe LoRA/hf4.x compatibility and BitNet packed-weight loading are separate issues." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer mapping helper failure is unrelated to the shape-mismatch index error." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative: it identifies the clear custom-model-init duplicate pair, treats the SIGLIP2 pair as likely rather than certain, and does not overclaim the many unrelated issues as related. The soft-edge rejections are also generally cautious and defensible from the titles alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 7, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44075", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44756", + "issue:44863", + "issue:44910", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45276", + "issue:45292", + "issue:45313", + "issue:45335", + "issue:45357", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45507", + "issue:45588", + "issue:45812", + "issue:46050", + "issue:46459", + "issue:46468", + "issue:46512", + "issue:46846", + "issue:46897", + "issue:46901" + ], + "soft_pairs": [ + "issue:43317|issue:44756", + "issue:38175|issue:39692", + "issue:43644|issue:44534", + "issue:41720|issue:42915", + "issue:42915|issue:44756", + "issue:42491|issue:45276", + "issue:44960|issue:44977", + "issue:44451|issue:44488", + "issue:44863|issue:46846", + "issue:46459|issue:46512", + "issue:45313|issue:45357", + "issue:45507|issue:46468", + "issue:43644|issue:43645", + "issue:39692|issue:43994", + "issue:43257|issue:43854", + "issue:44488|issue:44991", + "issue:43856|issue:43873", + "issue:43854|issue:43931", + "issue:42491|issue:45292", + "issue:43856|issue:44387", + "issue:43708|issue:44075", + "issue:42491|issue:45335", + "issue:44960|issue:46050", + "issue:44164|issue:44291", + "issue:46897|issue:46901", + "issue:45488|issue:45812", + "issue:45276|issue:45335", + "issue:45362|issue:45406", + "issue:44910|issue:45588", + "issue:44402|issue:44488", + "issue:39692|issue:44960" + ], + "prepared_review_unit_hash": "9d6f35dae5ae99e5e94f77991d3ea44700af6df3650b3f7bb90b5dc5d2af412d", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36359, + "estimated_input_tokens": 9090, + "estimated_eval_tokens": 18436 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "There is one clear duplicate pair in this cluster: the Transformers v5 non-persistent-buffer junk regression. The other soft pairs are conservative rejects because they share only broad themes (model family, loading, quantization, flash-attention, etc.) without the same concrete bug.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43644 is the earlier report of the same non-persistent-buffer junk regression; issue:44534 restates the same symptom with very similar wording.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the best representative issue in this set because it cleanly names the confirmed duplicate regression and is the earlier report in the pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Different problems: dequantized-model load/offload failure vs Strix Halo mmap-related OOM." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Both mention SigLIP2, but one is an inference/probability bug and the other is a docs example with model/processor mismatch and quantization errors." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same regression: Transformers v5 filling non-persistent buffers with junk; titles and symptom descriptions match closely." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Same broad Qwen3 area, but different concrete failures and code paths (cuda assert on auto device map vs FineGrainedFP8Config failure)." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Unrelated bugs: Qwen3 MoE FP8 config failure vs Strix Halo mmap/OOM workaround." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Different models and issues: LoRA/qwen3_moe compatibility across branches vs Gemma4 embedding resize propagation." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and failures: GLM5 issue vs Qwen3.5 flash-attention generation crash." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading failures, but they concern different checkpoints and there is not enough evidence they stem from the same code-path bug." + }, + { + "left": "issue:44863", + "right": "issue:46846", + "accept": false, + "reason": "NemotronH checkpoint loading issue vs Qwen3.5 TP-plan missing linear_attn entries; different causes." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Both involve dtype handling, but one is AutoModelForCausalLM loading of composite checkpoints and the other is from_config leaving config.dtype stale." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Different Qwen3.5 bugs: ZeRO-3 language_model weight loading vs save_pretrained visual-encoder key regression." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "Different models and failures; no shared code-path bug is evident." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Different regressions: buffer initialization bug vs Jupyter custom-model initialization breakage." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but one is a broken documentation example and the other is a bad AutoModel/pipeline output regression." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Different problems: Qwen3 MoE weight conversion with accelerate/deepspeed vs GLM-4.7-Flash loading in tests." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/loading regressions, but they affect different models and likely different tokenizer paths." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Different symptoms and mechanisms: Qwen3 MoE training memory inefficiency vs quantization offloading failure." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different load-time failures for different models: GLM-4.7-Flash test loading vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Unrelated bugs: cross-branch Qwen3 MoE LoRA compatibility vs generic resize_token_embeddings propagation issue." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both mention memory, but one is training inefficiency and the other is int4 quantization reserved-memory OOM." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint-resume math bug vs optimizer argument handling; different subsystems." + }, + { + "left": "issue:42491", + "right": "issue:45335", + "accept": false, + "reason": "Different model-specific bugs: qwen3_moe branch compatibility vs t5gemma embedding resize propagation." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "Different issues: GLM5 report vs quantized-model save/load breakage with sinq." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Both affect save/load, but one is extra_state handling and the other is init_empty_weights/_is_hf_initialized compatibility." + }, + { + "left": "issue:46897", + "right": "issue:46901", + "accept": false, + "reason": "Similar training-loss-shift pattern, but they are different model implementations and not clearly the same concrete bug." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Different tokenizer regressions in different model families; not the same underlying bug." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Similar API surface, but different model architectures and embedding components; insufficient evidence of one shared bug." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Different failures: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer in serve." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both touch flash_attention, but one is a packed-sequence/position_ids crash and the other is an s_aux=None AttributeError." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Different models and symptoms: vocab-size/token mismatch vs model loading failure." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "Unrelated model families and bugs; only superficial similarity in being model-specific regressions." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The core duplicate call on issue:43644 \u2194 issue:44534 is grounded, but the canonical_issue_reason overstates the evidence by calling 43644 the 'more complete' report. In the packet, 44534 has the longer body length, so completeness is not supported here. The rest of the soft-edge rejections read conservative enough, but this unsupported comparative claim means I can\u2019t accept the summary as written." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 8, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42947", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43645", + "issue:43646", + "issue:43708", + "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:44589", + "issue:44756", + "issue:44863", + "issue:44877", + "issue:44912", + "issue:44928", + "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:43425|issue:43957", + "issue:41720|issue:44155", + "issue:45663|issue:45759", + "issue:43404|issue:43454", + "issue:44291|issue:44756", + "issue:42947|issue:45704", + "issue:43873|issue:44387", + "issue:44079|issue:46829", + "issue:45663|issue:45859", + "issue:44292|issue:44912", + "issue:36010|issue:39404", + "issue:42947|issue:43381", + "issue:44155|issue:45823", + "issue:44387|issue:44756", + "issue:43957|issue:44756", + "issue:43335|issue:45663", + "issue:44863|issue:46468", + "issue:45276|issue:45292", + "issue:44479|issue:44877", + "issue:43381|issue:44928", + "issue:44164|issue:44756", + "issue:42796|issue:46326", + "issue:45588|issue:45663", + "issue:43854|issue:45216", + "issue:45684|issue:45698", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:42175|issue:43957", + "issue:44402|issue:44451", + "issue:43646|issue:43708", + "issue:44589|issue:44756" + ], + "prepared_review_unit_hash": "24780ce2d5060a6333a1dd156e862c60b0ded63649f46703cdccb8545ff9aae7", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36536, + "estimated_input_tokens": 9134, + "estimated_eval_tokens": 18524 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most candidate pairs are only loosely similar and should remain separate. The only clear duplicate chain is the `resize_token_embeddings` regression affecting output/input embeddings across model variants; everything else looks like distinct bugs.", + "confidence": 0.78, + "canonical_issue_reason": "Issue 45292 is the broadest, most generic report of the shared `resize_token_embeddings` bug, and it cleanly subsumes the model-specific variants in 45276 and 45335.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the best representative of the only convincing duplicate cluster here because it states the underlying bug without model-specific narrowing.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Both are loading-related, but one is a Torch compatibility failure and the other is a `torch.device(\"meta\")` model-loading regression; different root causes." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "CUDA assert during auto device mapping is unrelated to AudioFlamingo batched inference token/embedding leakage." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Gemma4 FSDP2 `KeyError` and Qwen multimodal `text_config` unwrapping are different code paths and different failures." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both mention lm_head tying, but they are model-specific reports with different affected architectures and likely different fixes." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "`init_empty_weights` argument handling and Strix Halo mmap/OOM are unrelated problems." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing with PEFT LoRA is a different bug from T5/apex RMSNorm memory leak." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both involve memory/quantization, but one is offloading behavior and the other is increased reserved CUDA memory under int4." + }, + { + "left": "issue:44079", + "right": "issue:46829", + "accept": false, + "reason": "ModelOutput key assignment logic is unrelated to VideoPrism's incorrect `hidden_states` field mapping." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Gemma4 FSDP2 attention state bug and Qwen3_5Moe CPU-offload incompatibility are separate issues." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "`init_empty_weights` loading error and MXFP4 quantization fallback are different model-loading regressions." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "Importing `GenerationMixin` and Whisper `return_language` pipeline behavior are unrelated." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "One is ineffective gradient checkpointing with LoRA; the other is checkpointing not allowed in eval mode." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo batching leak and Gemma4 PLE device mismatch are not the same underlying bug." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Int4 quantization OOM and Strix Halo mmap OOM are both memory-related but arise from different mechanisms." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading breakage and Strix Halo mmap disablement target different failure modes." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "SwitchTransformers sparse-layer config and Gemma4 FSDP2 attention state rebuilding are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint loading and the broken official Nemotron example are separate reports, not one concrete bug." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "Both describe the same `resize_token_embeddings` propagation failure; 45276 is the Gemma4-specific manifestation of the broader 45292 bug." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Qwen VL video-input regression and strict config loading for granite_speech are different subsystems and different bugs." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Eval-mode checkpointing restriction is unrelated to RLHF NaNs from 3D position_ids and SDPA fallback." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "`extra_state` save/load handling and Strix Halo mmap OOM do not share the same root cause." + }, + { + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer BPE-marker handling and `continue_final_message` behavior are both tokenizer-adjacent but distinct bugs." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Flash-attention `s_aux=None` crash and Gemma4 FSDP2 KeyError are separate implementation issues." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Unit-test loading of GLM-4.7-Flash and Qwen3.5 save_pretrained regression are unrelated." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Read-only permission propagation from `save_pretrained` is a different issue than loading the wrong custom module after saving." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Both are the same embedding-resize regression: 45292 is the generic output-embedding report and 45335 is the decoder-embedding variant." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Custom model initialization in notebooks is unrelated to `extra_state` save/from_pretrained handling." + }, + { + "left": "issue:42175", + "right": "issue:43957", + "accept": false, + "reason": "Missing TensorFlow in the torch extra and meta-device model-loading failures are unrelated." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Tokenizer vocab-size mismatch and ScandiBERT loading failure are different bugs." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization regression and `resume_from_checkpoint` max_steps calculation are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup failure and Strix Halo mmap/OOM are different low-level problems." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The only supported duplicate cluster is the resize_token_embeddings trio around 45292/45276/45335, and the other pair rejections are appropriately cautious." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 9, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:43299", + "issue:43425", + "issue:43452", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43645", + "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", + "issue:47050" + ], + "soft_pairs": [ + "issue:46894|issue:46901", + "issue:43425|issue:43828", + "issue:46315|issue:46491", + "issue:45812|issue:46315", + "issue:39404|issue:45701", + "issue:44530|issue:45084", + "issue:43611|issue:43645", + "issue:43425|issue:46869", + "issue:39404|issue:46665", + "issue:44683|issue:45823", + "issue:43425|issue:43716", + "issue:43425|issue:45823", + "issue:44479|issue:44805", + "issue:44589|issue:45084", + "issue:43901|issue:44509", + "issue:43906|issue:44530", + "issue:43994|issue:44079", + "issue:44265|issue:44898", + "issue:36032|issue:43452", + "issue:46050|issue:46326", + "issue:44164|issue:45759", + "issue:39404|issue:44912", + "issue:43906|issue:46326", + "issue:44671|issue:44991", + "issue:44991|issue:45081", + "issue:45823|issue:46899", + "issue:45479|issue:46894", + "issue:43856|issue:44756", + "issue:43299|issue:43931", + "issue:44164|issue:46821", + "issue:43632|issue:47050", + "issue:43596|issue:44683" + ], + "prepared_review_unit_hash": "2d55080040718aa84e31238e8de0321881a572fac0eb3ee92f0dbf0e3b2571d7", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37511, + "estimated_input_tokens": 9378, + "estimated_eval_tokens": 19012 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is dominated by unrelated Transformers v5 regression reports across different model families and subsystems. The soft pairs mostly share only broad symptom words (tokenization, dtype mismatch, loading, docs, or training loss), not a clearly identical bug or patch target, so they should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "Issue 45812 is the strongest representative because it is a concrete, high-signal v5 regression with a clear user-facing failure mode and broad impact within one subsystem (tokenization), unlike the more generic or model-specific reports elsewhere in the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45812 is the best overall issue to anchor this cluster: it is specific enough to be actionable, clearly describes a regression, and is more representative than vague umbrella reports or one-off model loading errors.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46894", + "right": "issue:46901", + "accept": false, + "reason": "Both describe a training-loss label-shift bug, but they are for different model families and lack evidence of the same shared implementation path." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "One is a Torch compatibility complaint, the other is a dtype mismatch in a specific model under autocast; the failure modes and code paths are different." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Performance regression in slow tokenizer conversion vs a CodeLlama encode/decode spacing bug; these are unrelated tokenizer issues." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Wrong token IDs for Granite models is a correctness regression; slow convert_tokens_to_ids is a performance regression, not the same bug." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline return_language and a generic tokenization-change report are not the same concrete failure." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and a compiler template-node error are different subsystems and different root causes." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both involve v5 loading behavior, but one is base_model_prefix loading and the other is custom model init in notebooks; no shared concrete bug is shown." + }, + { + "left": "issue:43425", + "right": "issue:46869", + "accept": false, + "reason": "Torch incompatibility and multi-GPU device-map failure are different issues; they only overlap at a high level." + }, + { + "left": "issue:39404", + "right": "issue:46665", + "accept": false, + "reason": "Whisper pipeline return_language and from_pretrained-inside-classes describe different call paths and different symptoms." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention failure on torch>=2.9 is unrelated to Gemma4 device mismatch in forward." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch version incompatibility is not the same as the Mistral-3 image preprocessor/model dtype mismatch." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "These are both broad compatibility complaints, but they target different models and distinct failures." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Both are regression reports, but one is a video-input model family regression and the other is a tensor-mask shape error; not the same bug." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup failure and compiler template-node errors are unrelated runtime/compiler problems." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but they discuss different pipeline/task deprecations and do not point to one underlying issue." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "An isolated reproduction of an older issue and a PagedAttentionCache crash on Qwen3.5 are different problems." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Wrong SigLIP2 outputs and incorrect ModelOutput key assignment are unrelated correctness bugs." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export failures via torch_compilable_check and Perceiver interpolation failures involve different code paths and different triggers." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Tokenizer add_special_tokens conflict and gguf_file from_pretrained breakage are separate loading issues." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantization save/load breakage and continue_final_message behavior in mistral_common tokenizers are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "extra_state save/load handling and from_config text_config unwrapping are different serialization/configuration bugs." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper language return and MXFP4 quantization fallback are unrelated regressions." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "These refer to different tokenizer behaviors and different model stacks; no shared defect is demonstrated." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT MLM prediction errors and tokenizer loading failures for EMBEDDIA/est-roberta are different issues." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-loading regressions, but one is a tokenizer file load issue and the other is a Mistral regex patch crash; not the same bug." + }, + { + "left": "issue:45823", + "right": "issue:46899", + "accept": false, + "reason": "Gemma4 device mismatch in forward and Gemma4 multimodal 4-bit dtype-cast breakage are distinct failures." + }, + { + "left": "issue:45479", + "right": "issue:46894", + "accept": false, + "reason": "Zero-loss classification bug and training-loss double-shift bug are both loss-related but not the same underlying defect." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory inefficiency and Strix Halo mmap OOM are separate performance/memory issues." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3VL MoE loading breakage and Qwen3-VL weight-shape mismatch are both model-loading issues, but with different concrete errors." + }, + { + "left": "issue:44164", + "right": "issue:46821", + "accept": false, + "reason": "extra_state save/load handling and TimesFM window_size AttributeError are unrelated issues." + }, + { + "left": "issue:43632", + "right": "issue:47050", + "accept": false, + "reason": "_is_hf_initialized flag regression and CanineModel fp16 ONNX export dtype mismatch are different bug classes." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Deepspeed ZeRO-3/BertModel index error and compiled flex_attention torch>=2.9 failure are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 10, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43317", + "issue:43404", + "issue:43421", + "issue:43452", + "issue:43482", + "issue:43541", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43746", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44462", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45479", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:46178", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46489", + "issue:46897" + ], + "soft_pairs": [ + "issue:45479|issue:46897", + "issue:44488|issue:44987", + "issue:44488|issue:45081", + "issue:43611|issue:43644", + "issue:42796|issue:43404", + "issue:44462|issue:46489", + "issue:36032|issue:42915", + "issue:46285|issue:46402", + "issue:43716|issue:45084", + "issue:39404|issue:44671", + "issue:44661|issue:44683", + "issue:43643|issue:44530", + "issue:44737|issue:45859", + "issue:42915|issue:44530", + "issue:44451|issue:44987", + "issue:40990|issue:43421", + "issue:43541|issue:45084", + "issue:34567|issue:40444", + "issue:43644|issue:43646", + "issue:40444|issue:46326", + "issue:43582|issue:43632", + "issue:42175|issue:46178", + "issue:44206|issue:44220", + "issue:36032|issue:43645", + "issue:44811|issue:45584", + "issue:43452|issue:43482", + "issue:44987|issue:44991", + "issue:45084|issue:45507", + "issue:43317|issue:43482", + "issue:43746|issue:46326" + ], + "prepared_review_unit_hash": "e128507d98f0b61cd5c14a15216d10a2ef8d5aa04560b0f3350029a19252e4e1", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36751, + "estimated_input_tokens": 9188, + "estimated_eval_tokens": 18632 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a grab bag of unrelated regression reports. The only plausible duplicate pair is the GGUF-loading regression on v5, where one issue is the generic breakage and the other is a model-specific manifestation. No PRs are present.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43452 is the broadest report of the GGUF loading regression, affecting both AutoTokenizer and AutoModelForCausalLM; it is the best umbrella issue for the related Qwen2.5-GGUF failure.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43452 is the strongest canonical issue candidate because it describes the underlying loading-path breakage in a model-agnostic way and can subsume the more specific GGUF symptom report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45479", + "right": "issue:46897", + "accept": false, + "reason": "Different problems: one is a generic zero-loss classification regression, the other is a Florence2 training-loss double-shift bug." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both are loading regressions, but they affect different models and likely different tokenizer/load paths." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer-loading failures: ScandiBERT load failure vs 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 junk in non-persistent buffers." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Unrelated issues: tokenizer BPE marker handling vs lm_head weight tying in Mistral3." + }, + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "Both involve tokenizers, but one is about AutoTokenizer ignoring tokenizer.json and the other is a DeepSeek-Coder output regression." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "Completely different areas: T5 tokenizer method conflict vs Qwen3Moe FP8 config failure." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "Different model-architecture regressions: CLIP structure mismatch vs Sam3 text encoder weights not loading." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Different codepaths: Mistral multimodal dtype mismatch vs a compilation/template-node error." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Different bugs in different models: Whisper pipeline return_language vs CamemBERT masked LM predictions." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "No shared underlying bug: model registration failure vs torch flex_attention compilation on newer torch." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "Trust-remote-code config field loss is unrelated to PagedAttentionCache's linear_attention crash." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "Both are device/offload-related, but they target different models and different failure mechanisms." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Same broad Qwen family, but distinct concrete failures: FP8 config vs PagedAttentionCache group-type crash." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Both are model-loading regressions, but they are different models and not clearly the same underlying defect." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity regression on a model is unrelated to runtime post-processor updates for special tokens." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Different failures: torch dynamo grouped_mm tracing error vs template-node compilation error." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "Trainer token-count tracking and multimodal iterable-dataset failures are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 custom-model regressions, but one concerns buffers and the other model initialization behavior." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "Different subsystems: multimodal fine-tuning with iterable datasets vs mistral_common tokenizer behavior." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "AppleSilicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Packaging/backend availability issues, but different backends and different failure points." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Related audio area, but one is a feature-extractor arg bug and the other is a lower-level fbank feature issue; not clearly the same defect." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer special-token conflict is unrelated to Jupyter custom model initialization." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both are Whisper special-token bugs, but one is decode-API behavior and the other is generation failure on empty transcription." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": true, + "reason": "Same underlying GGUF loading regression on v5: the generic `gguf_file` breakage matches the Qwen2.5-GGUF load failure." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Different offload/loading problems: dequantized model device_map handling vs Qwen2.5-GGUF version-specific load failure." + }, + { + "left": "issue:43746", + "right": "issue:46326", + "accept": false, + "reason": "Different model families and different load/runtime paths; no same-bug evidence." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative about the soft edges. The only accepted duplicate pair, issue:43452 vs issue:43482, is reasonably supported by the titles as a generic GGUF loading regression versus a model-specific manifestation. The rest of the verdicts appropriately avoid over-claiming similarity, and there are no PR pairs to assess." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 11, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28530", + "issue:34567", + "issue:36032", + "issue:41720", + "issue:42915", + "issue:43089", + "issue:43208", + "issue:43262", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43531", + "issue:43540", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43792", + "issue:43819", + "issue:43883", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44849", + "issue:44877", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45987", + "issue:46661", + "issue:46821", + "issue:46869", + "issue:46899", + "issue:47013" + ], + "soft_pairs": [ + "issue:43262|issue:43377", + "issue:43208|issue:47013", + "issue:43883|issue:44488", + "issue:43792|issue:46869", + "issue:44403|issue:45081", + "issue:43646|issue:44534", + "issue:43645|issue:46661", + "issue:42915|issue:45084", + "issue:44849|issue:45325", + "issue:43377|issue:43819", + "issue:43404|issue:44466", + "issue:44987|issue:45081", + "issue:28530|issue:34567", + "issue:43994|issue:44610", + "issue:44464|issue:45987", + "issue:42915|issue:44291", + "issue:44451|issue:45081", + "issue:36032|issue:43742", + "issue:44008|issue:46899", + "issue:44062|issue:44987", + "issue:44265|issue:44610", + "issue:43322|issue:43540", + "issue:42915|issue:44589", + "issue:43632|issue:46821", + "issue:44479|issue:45325", + "issue:43883|issue:44991", + "issue:43089|issue:45987", + "issue:43531|issue:44877", + "issue:36032|issue:41720" + ], + "prepared_review_unit_hash": "53e3b227ebbee2bae89fb56aeb9a9bcd75629b08d7feb5ed72abb82bc1771b3e", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36004, + "estimated_input_tokens": 9001, + "estimated_eval_tokens": 18258 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is a collection of mostly unrelated Transformers bug reports; the soft pairs are mostly same-surface-area but not the same underlying defect, so they should stay separate.", + "confidence": 0.65, + "canonical_issue_reason": "Issue 43646 is the broadest, most central regression-style report in the set: Transformers v5 breaking custom model initialization. It is a reasonable anchor for the cluster\u2019s general v5/model-init breakage theme.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43646 is the best standalone representative because it describes a general, high-impact API break rather than a model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Both are audio-related, but one is a chat-template sampling-rate default bug and the other is a batching/padding-mask encoder mismatch." + }, + { + "left": "issue:43208", + "right": "issue:47013", + "accept": false, + "reason": "Both involve xLSTM training, but they describe different failures: multiple bugs for small models vs a train-mode unpacking crash." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading errors, but they affect different models and different code paths." + }, + { + "left": "issue:43792", + "right": "issue:46869", + "accept": false, + "reason": "No concrete overlap beyond generic runtime failure; different models and different device-mapping behavior." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "One is about noisy loading messages, the other is a tokenizer regex patch crash; not the same bug." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but one is custom model initialization and the other is non-persistent buffer serialization corruption." + }, + { + "left": "issue:43645", + "right": "issue:46661", + "accept": false, + "reason": "Both concern custom/remote-code model initialization, but the failures are different: notebook-defined custom models vs missing config_class on remote-code models." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Different subsystems: fine-grained FP8 Qwen3Moe failure vs a template compilation TypeError." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen3.x multimodal regressions, but one affects hidden states and the other rope-index scaling for still images." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Both involve audio codecs, but one is padding-mask batching and the other is DAC latent decoding mismatch." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied lm_head weights, but one is an actual model tying bug and the other is device-dependent serialization inconsistency." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Different loading failures: one is a version-specific model load failure, the other is a Mistral tokenizer regex patch crash." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Completely unrelated training-state bugs: early stopping metric selection vs token-count tracking." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Different components and symptoms: bad AutoModel/pipeline outputs vs processor image-size mismatch." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "One is chunked-generation inconsistency with compiled forward; the other is a cache-length return-type contract issue." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Both touch loading/initialization, but one is Qwen3Moe FP8 and the other is init_empty_weights/_is_hf_initialized compatibility." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer-loading failures; no evidence they share the same root cause." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Different tokenizer/model issues: T5 tokenizer API conflict vs a key error loading a specific model." + }, + { + "left": "issue:44008", + "right": "issue:46899", + "accept": false, + "reason": "Both are Gemma multimodal issues, but one is a forward variable collision and the other is unsafe dtype casting under 4-bit quantization." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer construction bug vs model-loading regression; not the same underlying defect." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "Export/compilation failure from torch_compilable_check is unrelated to OmDet processor image-size mismatch." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Different multimodal failures: Llava Next segfault vs Qwen3OmniMoe video-processing ValueError." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Different low-level failures: FP8 config handling vs missing Float8 storage type." + }, + { + "left": "issue:43632", + "right": "issue:46821", + "accept": false, + "reason": "Both are v5-era regressions, but one is the _is_hf_initialized flag and the other is a TimesFM window_size attribute error." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Related Qwen VL family, but one is a video-input regression and the other is still-image position-id scaling." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Both are loading/tokenizer-adjacent, but they are different models and different error modes." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation performance overhead is not the same as an incorrect cache-length return type." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "No shared bug mechanism; sliding-window behavior and strict config validation are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer method conflict and multi-GPU Qwen device-map crash are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 12, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43296", + "issue:43317", + "issue:43381", + "issue:43421", + "issue:43425", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43646", + "issue:43716", + "issue:43746", + "issue:43792", + "issue:43856", + "issue:44079", + "issue:44220", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44991", + "issue:45081", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:46315", + "issue:46491", + "issue:46710", + "issue:46821", + "issue:46899" + ], + "soft_pairs": [ + "issue:45216|issue:45357", + "issue:43425|issue:44403", + "issue:44991|issue:46315", + "issue:40444|issue:45325", + "issue:43296|issue:43531", + "issue:43716|issue:46899", + "issue:43611|issue:44534", + "issue:44220|issue:44683", + "issue:44403|issue:44991", + "issue:45812|issue:46491", + "issue:42796|issue:43746", + "issue:43381|issue:45923", + "issue:43632|issue:44756", + "issue:33666|issue:40444", + "issue:43792|issue:44589", + "issue:40990|issue:44811", + "issue:42796|issue:46710", + "issue:44291|issue:44589", + "issue:35707|issue:40444", + "issue:43482|issue:43525", + "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:44079|issue:44534", + "issue:43317|issue:43856", + "issue:44462|issue:45812", + "issue:43296|issue:46821", + "issue:43611|issue:43646" + ], + "prepared_review_unit_hash": "0d7a09e13ff5d474683eff1edce4427373aeec342184cc8b1b483404b4f53841", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36961, + "estimated_input_tokens": 9241, + "estimated_eval_tokens": 18738 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated bug reports. The only clear duplicate-like pair is the Qwen3.5 save_pretrained regression, where the later report appears to be the same checkpoint/key corruption issue resurfacing.", + "confidence": 0.74, + "canonical_issue_reason": "issue:45216 is the clearest concrete regression and has the strongest duplicate-style overlap with issue:45357 about Qwen3.5 save_pretrained producing incorrect checkpoint contents.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the best representative issue because it is specific, regression-focused, and has the clearest evidence of a same-bug follow-up in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same Qwen3.5 save_pretrained regression; both report incorrect saved checkpoint/visual encoder keys in v5.x." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch 2.10 compatibility failure vs noisy load logging; different bugs." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer loading failure vs slow convert_tokens_to_ids regression; related area but different concrete problems." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Both involve Qwen2.5-VL, but one is dataset/mult-image finetuning failure and the other is a rope/indexing bug." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL load failure vs Qwen3-MoE sliding_window issue; unrelated." + }, + { + "left": "issue:43716", + "right": "issue:46899", + "accept": false, + "reason": "Different multimodal dtype issues on different models and code paths; not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions during model loading, but one is base_model_prefix handling and the other is buffer initialization; not the same defect." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Audio feature extraction bug vs compiled flex_attention torch-version issue; unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Generic load noise vs tokenizer loading failure; different symptoms and fixes." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but one is Granite token IDs and the other is CodeLlama round-trip spacing; not the same issue." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Tokenizer BPE-marker handling vs PEFT/local-checkpoint loading for GraniteSpeech; unrelated." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Eval-mode gradient checkpointing limitation vs missing supports_gradient_checkpointing flag; different code paths." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "HF initialization flag regression vs mmap/OOM workaround; unrelated." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Qwen2-VL multi-GPU training vs Qwen2.5-VL multi-image dataset failure; both VL training, but not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper run failure vs Float8 storage lookup error; different failures." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity regression on GPT-OSS vs Whisper batch_decode special-token handling; unrelated." + }, + { + "left": "issue:42796", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer marker bug vs accuracy regression on a different model; not the same change." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized argument issue vs missing Float8 storage object; different bugs." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values vs multi-image finetuning failure; unrelated." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Qwen2.5-GGUF loading failure vs Llama4Config pad_token_id attribute error; different models and failures." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment bug vs Gemma4 FSDP2 shared_kv_states KeyError; unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug vs init_empty_weights argument bug; not the same problem." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Runtime special-token post-processor update bug vs add_special_tokens BOS/EOS regression; similar area but different concrete failure." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Auto device mapping cudaErrorAssert vs Mistral regex patch tokenizer crash; unrelated." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Missing TensorFlow backend in torch-only install vs tensorboard-without-TF backend error; different packaging paths." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "GGUF loading failure vs CLIPTokenizer max_length default bug; unrelated." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key bookkeeping vs non-persistent buffer junk values; different internals." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "device_map=auto dequantized-model offload failure vs Qwen3 MoE memory usage regression; not the same bug." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json vs wrong Granite token IDs; different tokenizer regressions." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "PaddleOCR-VL load failure vs TimesFM window_size AttributeError; unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both are Transformers v5 loading regressions, but one is base_model_prefix handling and the other is custom model initialization; too broad to merge as one bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The accepted 45216\u219445357 pair is well supported by the titles, while the rejected pairs are mostly distinct bugs with appropriately cautious reasoning." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 13, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43262", + "issue:43296", + "issue:43377", + "issue:43454", + "issue:43482", + "issue:43493", + "issue:43550", + "issue:43582", + "issue:43606", + "issue:43644", + "issue:43701", + "issue:43746", + "issue:43818", + "issue:43828", + "issue:43873", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44479", + "issue:44509", + "issue:44829", + "issue:44877", + "issue:45072", + "issue:45325", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45491", + "issue:45584", + "issue:45701", + "issue:45823", + "issue:45834", + "issue:46402", + "issue:46821", + "issue:46909" + ], + "soft_pairs": [ + "issue:44509|issue:46909", + "issue:42175|issue:43296", + "issue:43377|issue:45491", + "issue:43644|issue:44079", + "issue:40444|issue:45381", + "issue:43818|issue:46402", + "issue:43818|issue:45357", + "issue:45325|issue:45834", + "issue:44265|issue:45072", + "issue:43493|issue:43994", + "issue:45325|issue:45381", + "issue:44829|issue:45479", + "issue:43454|issue:44466", + "issue:41553|issue:43582", + "issue:42915|issue:43482", + "issue:44291|issue:44877", + "issue:43701|issue:45325", + "issue:42796|issue:45381", + "issue:43873|issue:45701", + "issue:40990|issue:45584", + "issue:43606|issue:44265", + "issue:44479|issue:45823", + "issue:43262|issue:43746", + "issue:43828|issue:46821", + "issue:35707|issue:41720", + "issue:34634|issue:35707", + "issue:39692|issue:40444", + "issue:44186|issue:44265", + "issue:41720|issue:45823", + "issue:43454|issue:43550" + ], + "prepared_review_unit_hash": "a6239947f8ec760c93c054f340644e850f3b1a16fd364098623e06130f1b1a93", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36926, + "estimated_input_tokens": 9232, + "estimated_eval_tokens": 18720 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous, but two soft-related subthreads stand out: the SigLIP2 implementation/output regression pair and the Qwen2.5-VL visual position-id/rope-index regression pair. Those matches look like the closest counterparts in the packet, but they should be treated as likely related rather than confirmed duplicate identities.", + "confidence": 0.78, + "canonical_issue_reason": "No single canonical issue cleanly represents the whole cluster because the items span unrelated bugs. If a representative issue is needed, issue:43493 is the best closest counterpart for the SigLIP2 subthread because it describes the root implementation discrepancy behind the downstream bad outputs seen in 43994.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43493 is the strongest issue anchor overall: it is concrete and root-cause oriented, and it is the best closest counterpart for the SigLIP2-related reports. issue:45325 is the other strong candidate for the Qwen2.5-VL position-id regression subthread, but it is a separate thread.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44509", + "right": "issue:46909", + "accept": false, + "reason": "Both are documentation bugs, but they concern different docs/examples and different failure modes; not the same underlying issue." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Unrelated problems: backend packaging vs a PaddleOCR-VL load failure." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both mention batching/padding, but they are different models and different bugs." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Different code paths: non-persistent buffer init vs ModelOutput key assignment." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Same broad model family, but one is finetuning with multiple images and the other is a video position-id regression; not the same concrete bug." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Different models and failures: video_tower architecture issue vs text-encoder weight loading." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Different models and different save/load bugs; only a broad visual-component theme overlaps." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Both involve position/index logic, but the model, modality, and failure are different." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "Different failures in different subsystems: export handling vs dtype mismatch in inference." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 correctness thread: 43493 describes the HF-vs-JAX implementation discrepancy, and 43994 shows the downstream nonsensical outputs from that bug." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": true, + "reason": "Likely the same Qwen2.5-VL visual position-id / rope-index bug affecting image/video inputs, though the packet only supports a soft duplicate read." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Different training pathologies with different triggers; not the same bug." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied lm_head weights, but one is a model-specific tie failure and the other is a serialization inconsistency." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Completely unrelated: tokenizer error messaging vs Apple Silicon allocator TypeError." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Different quantization/load paths and different models; only a generic loading failure overlaps." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "Different root causes: init_empty_weights argument mismatch vs strict config loading." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to visual position-id computation." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video position ids are different bugs." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantized offloading vs version-dependent tokenization are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "Different models and symptoms: perplexity evaluation vs empty-transcription generation failure." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "Different runtime failures: CPU offload device mismatch vs export-time torch_compilable_check issues." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Different models and different device-map problems; too broad to be the same bug." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and GraniteSpeech checkpoint loading are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:46821", + "accept": false, + "reason": "Different failures: autocast dtype mismatch vs missing window_size attribute." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Different model behaviors and different failure modes; no shared underlying bug." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice preset and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "Both are multimodal/model-usage issues, but they are about different models and distinct bugs." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Tokenizer padding/truncation crashes are unrelated to torch.export exportability." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Different models and different device-mismatch mechanisms; not the same bug." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "Tied-weight generation bug and torch.compile SDPA failure are unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level read is mostly grounded, but a few phrases overstate the evidence. In particular, calling issue:43994 a \"true-duplicate counterpart\" of issue:43493 is stronger than the packet supports, since the available material only shows a similar SigLIP2 root-cause/symptom relationship, not an explicit duplicate link. Likewise, the 45325\u219445381 match is plausible but should be described more cautiously as a likely or soft duplicate, not as confirmed same-bug identity. Please soften those claims and keep the anchor language to \"closest counterpart\" / \"likely related\" rather than definitive duplicate wording." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 14, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:26413", + "issue:28530", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42796", + "issue:42915", + "issue:42947", + "issue:43065", + "issue:43257", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43450", + "issue:43454", + "issue:43572", + "issue:43606", + "issue:43720", + "issue:43749", + "issue:43819", + "issue:43873", + "issue:43881", + "issue:44387", + "issue:44466", + "issue:44488", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45357", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45920", + "issue:45923", + "issue:46050", + "issue:46209", + "issue:46305", + "issue:46315", + "issue:46402", + "issue:46821" + ], + "soft_pairs": [ + "issue:43606|issue:43819", + "issue:43257|issue:45923", + "issue:42796|issue:46209", + "issue:44568|issue:45920", + "issue:43377|issue:44610", + "issue:45823|issue:46821", + "issue:43454|issue:45072", + "issue:46305|issue:46315", + "issue:40990|issue:42947", + "issue:42617|issue:42915", + "issue:43377|issue:44898", + "issue:43572|issue:45759", + "issue:43317|issue:43873", + "issue:43454|issue:43720", + "issue:43749|issue:44534", + "issue:43881|issue:44488", + "issue:43322|issue:46050", + "issue:41720|issue:43322", + "issue:36331|issue:42617", + "issue:41720|issue:44977", + "issue:44387|issue:45704", + "issue:36331|issue:42915", + "issue:44877|issue:44991", + "issue:40990|issue:45701", + "issue:44466|issue:45357", + "issue:43450|issue:45701", + "issue:43065|issue:46402", + "issue:41720|issue:42617", + "issue:26413|issue:28530", + "issue:43873|issue:44756" + ], + "prepared_review_unit_hash": "ee4f1fb3c5d26402a76cf03d9b138ca90310095bef26203c36aae59ba332ef40", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36636, + "estimated_input_tokens": 9159, + "estimated_eval_tokens": 18574 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly distinct transformer/model bugs with only superficial overlap (often the same broad area like offloading, tokenization, or saving/loading). I found no soft pair that is clearly the same underlying bug/change, so all pairwise duplicate links should be rejected.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43873 is the broadest and most centrally referenced issue in the set, with a general offloading/quantization regression theme and the highest inbound-reference count among the candidates.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best single issue candidate because it is broad, well-discussed, and likely to capture related offloading/quantization regressions better than the narrower model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Different models and different failures: CPU-offload device mismatch vs DAC latent/forward mismatch." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Qwen3 MoE loading/DeepSpeed issue versus a trust_remote_code gradient-checkpointing flag omission." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE-marker handling bug is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both are tokenizer regressions, but one is a specific BOS/EOS special-token bug and the other is a broad wrong-ID regression across several models." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Both involve batched inputs, but they concern different components and different concrete bugs." + }, + { + "left": "issue:45823", + "right": "issue:46821", + "accept": false, + "reason": "Device-mapping forward crash versus a TimesFM window_size AttributeError; unrelated code paths." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "lm_head weight tying/serialization problem is not the same as a bfloat16 dtype mismatch." + }, + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "Sliding-window mask off-by-one is unrelated to tokenizer conversion performance." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "High perplexity on a model is not the same bug as PEFT LoRA gradient-checkpointing ineffectiveness." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel.py runtime failure is unrelated to Qwen3 MoE fine-grained FP8 loading." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Different vision/input bugs: padding-mask batch mismatch versus incorrect resolution handling in Perceiver." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_idx in one config is unrelated to unwrapping text_config for composite Qwen multimodal configs." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both mention offloading, but one is dequantized-model loading and the other is broader quantization/offload behavior; not the same concrete bug." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "Different serialization/loading issues: tied lm_head weights versus packed BitNet weights not being unpacked." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM-efficient loading failure is unrelated to non-persistent buffers being filled with junk." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Two model-loading failures, but for different models and different root causes." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Segfault when loading Llava Next is unrelated to sinq quantization save/load corruption." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Both are severe load-time failures, but for different models and failure modes (cudaErrorAssert vs segfault)." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer signature change bug is unrelated to a parallel script runtime failure." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Same model family, but different concrete bugs: device-mapping assert versus flash-attention generation failure." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Memory pressure under int4 quantization is not the same as a T5 Apex RMSNorm memory leak." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "Unexpected compute_loss keyword argument is unrelated to Qwen3 MoE FP8 loading." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config loading failure and tokenizer-loading regression are different code paths." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "Perplexity regression is unrelated to tokenization changing across versions." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both are save_pretrained regressions, but they affect different tensors/keys and likely different fixes." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Incorrect video batch shape is unrelated to tokenization changes." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "Sam3 pixel-decoder dummy Conv2d is unrelated to Sam3 text-encoder weight loading." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Different systems and failure modes: auto device-mapping assert versus 3d_parallel script failure." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Checkpoint resume bug is unrelated to early-stopping metric resolution." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Quantization/offloading regression is unrelated to disabling mmap to avoid Strix Halo OOM." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 15, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28530", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43317", + "issue:43404", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43646", + "issue:43708", + "issue:43720", + "issue:43881", + "issue:43883", + "issue:44008", + "issue:44186", + "issue:44423", + "issue:44451", + "issue:44464", + "issue:44610", + "issue:44671", + "issue:44734", + "issue:44743", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45632", + "issue:45684", + "issue:45685", + "issue:45701", + "issue:45823", + "issue:46291", + "issue:46612", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:46869", + "issue:47020" + ], + "soft_pairs": [ + "issue:44186|issue:45072", + "issue:46612|issue:46858", + "issue:43577|issue:43708", + "issue:43550|issue:43720", + "issue:43295|issue:45701", + "issue:45685|issue:46821", + "issue:43646|issue:46665", + "issue:44451|issue:47020", + "issue:43881|issue:44991", + "issue:43526|issue:43883", + "issue:44186|issue:44610", + "issue:43550|issue:43606", + "issue:36010|issue:42617", + "issue:43404|issue:45357", + "issue:42915|issue:43317", + "issue:44610|issue:44898", + "issue:46821|issue:46869", + "issue:44734|issue:46858", + "issue:44423|issue:45406", + "issue:44008|issue:45823", + "issue:41720|issue:45161", + "issue:28530|issue:33666", + "issue:43531|issue:43572", + "issue:44464|issue:46858", + "issue:42915|issue:46291", + "issue:36331|issue:41720", + "issue:44671|issue:44743", + "issue:45632|issue:45684" + ], + "prepared_review_unit_hash": "efff030601e9dd4dcb5469a3b985c8c31ff469f2ba2f3d4ecc012d8c99e419d1", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36301, + "estimated_input_tokens": 9076, + "estimated_eval_tokens": 18408 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly unrelated bug reports across training, tokenizers, multimodal processors, generation cache logic, and model loading. The soft overlaps are only superficial; none look like the same underlying bug or change.", + "confidence": 0.89, + "canonical_issue_reason": "issue:46858 is the most concrete, self-contained bug report in the set and is closest to the cache/generation cluster, but there is no true duplicate hub here.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46858 has the clearest symptom, reproduction, and code-path specificity among these unrelated reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Different subsystems and failures: tokenizer/NER padding crash vs dtype mismatch in inference." + }, + { + "left": "issue:46612", + "right": "issue:46858", + "accept": false, + "reason": "Both involve generation cache code, but the bugs are different: skipped beam cache reorder vs static-cache seq_length arithmetic on non-CUDA backends." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Unrelated issues: BLIP2 dtype loading vs Trainer resume max_steps calculation." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "Different concrete failures: SDPA with torch.compile vs packed weights not unpacked during accelerate loading." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Both are regressions around loading/tokenization behavior, but one is processor.tokenizer access and the other is version-dependent tokenization differences." + }, + { + "left": "issue:45685", + "right": "issue:46821", + "accept": false, + "reason": "No shared bug: MPS histogram support in MoE vs TimesFM window_size attribute access." + }, + { + "left": "issue:43646", + "right": "issue:46665", + "accept": false, + "reason": "Both mention custom loading flows, but one is Transformers 5 custom model init breakage and the other is from_pretrained inside Python classes." + }, + { + "left": "issue:44451", + "right": "issue:47020", + "accept": false, + "reason": "Tokenizer loading failures, but for different root causes and model families; not the same underlying issue." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Different failures: glm-4v-9b load error vs tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "Completely different bugs: BEiT reduce_labels output vs missing all_tied_weights_keys on MolmoForCausalLM." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Both are preprocessing/processor issues, but one is text/tokenization padding and the other is image size mismatch." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Different code paths: torch.compile SDPA on Bamba vs CPU offload device mismatch in Bark." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "No meaningful overlap: GenerationMixin import error vs a 3d_parallel.py runtime failure." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Different model-save/load regressions: tied lm_head weights vs incorrect visual encoder keys." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Both are model-loading/runtime failures, but the causes differ: FP8 config failure vs dequantized model offload failure." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Different model families and failure modes: OmDet-Turbo image-size mismatch vs Perceiver positional encoding failure." + }, + { + "left": "issue:46821", + "right": "issue:46869", + "accept": false, + "reason": "Unrelated: TimesFM window_size AttributeError vs YOLO multi-GPU device-map loading." + }, + { + "left": "issue:44734", + "right": "issue:46858", + "accept": false, + "reason": "Both touch cache behavior, but one is KV-cache continuation indexing in serve and the other is seq_length device arithmetic in static cache." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both are serve/multimodal crashes, but the underlying bugs differ: string-to tensor conversion path vs Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Same broad Gemma area, but different bugs: variable name collision in Gemma3n vs device mismatch in Gemma4 PLE." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Different model-specific runtime issues: Qwen3 auto device-map assertion vs GPT-OSS MoE tensor-parallel behavior." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Training early-stopping configuration bug vs Qwen2-VL multi-GPU training request." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs StableLmConfig missing pad_token_idx." + }, + { + "left": "issue:44464", + "right": "issue:46858", + "accept": false, + "reason": "Both involve compiled generation, but one is inconsistent chunked outputs and the other is static-cache sequence-length arithmetic." + }, + { + "left": "issue:42915", + "right": "issue:46291", + "accept": false, + "reason": "Completely unrelated: FP8 model failure vs kernels package import/version requirement." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Different layers and symptoms: Trainer API change in compute_loss vs a CUDA assert in Qwen3 auto device mapping." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "Distinct model behaviors: CamemBERT masked-LM predictions vs Qwen3.5 recurrent state reset with cache." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Different save_pretrained problems: cache-path collision vs read-only permission propagation." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 16, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43881", + "issue:44186", + "issue:44361", + "issue:44387", + "issue:44492", + "issue:44509", + "issue:44534", + "issue:44589", + "issue:44913", + "issue:45059", + "issue:45406", + "issue:45464", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:47020" + ], + "soft_pairs": [ + "issue:44186|issue:44361", + "issue:43425|issue:46821", + "issue:43316|issue:44913", + "issue:28530|issue:30064", + "issue:42915|issue:43322", + "issue:33357|issue:35707", + "issue:43540|issue:43782", + "issue:29159|issue:36032", + "issue:30064|issue:34634", + "issue:43873|issue:45704", + "issue:44387|issue:45701", + "issue:44534|issue:46665", + "issue:43531|issue:45759", + "issue:43550|issue:43653", + "issue:43452|issue:47020", + "issue:44492|issue:44509", + "issue:41720|issue:43746", + "issue:43606|issue:43720", + "issue:42490|issue:43329", + "issue:45406|issue:45464", + "issue:43388|issue:45059", + "issue:34567|issue:35707", + "issue:43531|issue:46821", + "issue:43606|issue:43653", + "issue:43526|issue:43881", + "issue:34634|issue:41720", + "issue:43540|issue:46858", + "issue:43531|issue:44589", + "issue:45759|issue:45859" + ], + "prepared_review_unit_hash": "02c50ccfef6a82d78ac5b8adbceb0bdd4b5dadd32fa0f10cfc4e89a20c771014", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35785, + "estimated_input_tokens": 8947, + "estimated_eval_tokens": 18150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose collection of unrelated Transformers issues; the soft-similarity pairs share vocabulary but not the same underlying bug, so none should be merged as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Both are tokenizer/task crashes, but they affect different models and failure modes (LayoutLMv2 NER/padding vs MLuke task AttributeError)." + }, + { + "left": "issue:43425", + "right": "issue:46821", + "accept": false, + "reason": "Unrelated problems: Torch 2.10 compatibility vs TimesFM window_size AttributeError." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Both mention config API behavior, but one is Gemma3TextConfig parity and the other is GPTNeoX rotary_pct reload persistence." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Early-stopping metric selection and image processor void segmentation maps are different subsystems and bugs." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Different model-loading failures for different models; no shared concrete code path." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Different Qwen model families and different from_pretrained/video-loading errors." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Tokenizer inconsistencies are too broad here; one is slow-vs-fast parity, the other is a T5 add_special_tokens name conflict." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Image segmentation-map processing and Bark voice_preset handling are unrelated processor bugs." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantization/offloading behavior and T5 apex FusedRMSNorm memory leak are different problems." + }, + { + "left": "issue:44387", + "right": "issue:45701", + "accept": false, + "reason": "Int4 CUDA OOM and tokenization changes across versions are not the same issue." + }, + { + "left": "issue:44534", + "right": "issue:46665", + "accept": false, + "reason": "Non-persistent buffer initialization junk and from_pretrained-in-python-class behavior are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Both involve Qwen configs, but one is sliding_window on Qwen3-MoE and the other is text_config unwrapping for multimodal configs." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "Different components and failures: torch.compile/SDPA vs tokenizer special-token registration." + }, + { + "left": "issue:43452", + "right": "issue:47020", + "accept": false, + "reason": "Both are loading/tokenizer-related, but one is GGUF auto-class loading and the other is XLMRobertaTokenizer passing the wrong type to Unigram." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A cache-strategy typo and deprecated docs for removed pipeline tasks are not the same change or bug." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different models and loading paths: Qwen3 auto device mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve accelerate/offload-like loading, but the models and root causes differ (device mismatch vs packed weight unpacking)." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Local-path import warning and multimodal token counting video-branch undefined variables are unrelated." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming chat/completions failures do not share a concrete bug." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Metric gathering on tuple labels and SAM3 text/bounding-box behavior are different code paths." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "Trainer token-count state tracking and progressive generation are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:46821", + "accept": false, + "reason": "Different models and different bugs; no shared underlying fix." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "CPU offload device mismatch and BigBirdTokenizer special-token decode behavior are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and glm-4v-9b loading failure are unrelated." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset and Qwen3 auto device mapping are different subsystems and errors." + }, + { + "left": "issue:43540", + "right": "issue:46858", + "accept": false, + "reason": "Qwen3OmniMoe video-input validation and static-cache slice arithmetic on non-CUDA backends are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and Float8 storage lookup failure are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Different Qwen multimodal config/model issues: from_config text_config unwrapping vs CPU-offload rotary embedding." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 17, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:41720", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43493", + "issue:43644", + "issue:43653", + "issue:43746", + "issue:43782", + "issue:43844", + "issue:43873", + "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:45381", + "issue:45479", + "issue:45704", + "issue:45759", + "issue:45923", + "issue:46032", + "issue:46050", + "issue:46402", + "issue:46463", + "issue:46655", + "issue:46665", + "issue:46693", + "issue:47050" + ], + "soft_pairs": [ + "issue:45357|issue:46402", + "issue:44451|issue:44991", + "issue:44077|issue:44625", + "issue:43644|issue:46665", + "issue:44987|issue:45042", + "issue:44062|issue:44451", + "issue:44898|issue:45072", + "issue:44625|issue:45479", + "issue:43381|issue:43844", + "issue:39692|issue:44610", + "issue:44368|issue:45357", + "issue:43746|issue:46050", + "issue:41720|issue:45704", + "issue:46655|issue:46693", + "issue:43381|issue:45704", + "issue:44918|issue:44936", + "issue:43421|issue:43927", + "issue:42886|issue:44843", + "issue:44514|issue:44545", + "issue:45704|issue:45923", + "issue:45759|issue:45923", + "issue:43746|issue:43873", + "issue:43994|issue:44671", + "issue:46032|issue:46463", + "issue:39692|issue:43493", + "issue:43421|issue:43653", + "issue:30064|issue:36010", + "issue:43782|issue:45381", + "issue:44464|issue:47050" + ], + "prepared_review_unit_hash": "9497bbcd809b970084854bfed706d46e9e1e1d20476042f3bb6e810b4ccee9ab", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36683, + "estimated_input_tokens": 9171, + "estimated_eval_tokens": 18598 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous and contains only a couple of true duplicates. The strongest matches are the offline tokenizer/cache failure pair (42886 \u2194 44843) and the batched Qwen2_5_VLProcessor crash pair (44514 \u2194 44545).", + "confidence": 0.96, + "canonical_issue_reason": "The cluster does not have one unifying issue, but issue 44545 is the cleanest canonical anchor among the true duplicates because it is an exact title match with 44514 and describes a focused, reproducible processor crash.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44545 is the best issue-level representative because it forms an exact duplicate pair with 44514; the rest of the items are largely unrelated bug reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Both are model-loading regressions, but they affect different models and different failure modes: visual encoder keys vs missing text encoder weights." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/load failures, but they target different models and likely different code paths; not the same underlying bug." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "These are unrelated: one is about patchtsmixer post_init policy, the other about Qwen3.5 num_labels propagation." + }, + { + "left": "issue:43644", + "right": "issue:46665", + "accept": false, + "reason": "Different problems: non-persistent buffer initialization junk vs inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "One is a model load failure for a specific repo; the other is an image-processor dependency regression requiring torchvision." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken argument duplication is a tokenizer construction bug, while the other is a model-specific load failure." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Perceiver interpolation at non-default resolution and bfloat16 dtype mismatches are unrelated code-path failures." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Both touch classification config, but one is missing num_labels propagation and the other is a degenerate loss bug." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Gradient checkpointing in eval mode and ZeRO-3 gradient explosion are different issues with different causes." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "A broken documentation example for SigLIP2 is unrelated to OmDet-Turbo producing the wrong processor image size." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "A tie_word_embeddings warning in fine-tuning is unrelated to incorrect visual encoder keys saved by Qwen3.5." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Different load/save bugs: PEFT adapters not loading from local checkpoints vs quantized sinq models breaking after save/load." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "Qwen3 auto device mapping cuda assert and T5 apex FusedRMSNorm memory leak are unrelated." + }, + { + "left": "issue:46655", + "right": "issue:46693", + "accept": false, + "reason": "Incorrect video padding in SmolVLM is unrelated to a FA2 long-context performance regression." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode restrictions have nothing to do with T5's apex memory leak." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Unpacking Qwen3.5 input embeddings in TRL SFT and trainer.evaluate failing after train are different trainer behaviors." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Special-token post-processor refresh and DiaConfig custom token-ID persistence are related broadly, but not the same concrete bug." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same offline-mode tokenizer/cache failure: both describe HF_HUB_OFFLINE breaking AutoTokenizer because an unconditional hub/model_info lookup is performed." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Exact duplicate: same title, same crash location, same batched-input/padding=False failure in Qwen2_5_VLProcessor.apply_chat_template." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Both involve special-token handling, but one is backend post-processor synchronization and the other is BigBirdTokenizer mask-token registration." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image Processor void-segmentation-map handling is unrelated to the GenerationMixin import error." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen multimodal bugs: weight_only loading failure vs incorrect vision_position_ids in video input." + }, + { + "left": "issue:44464", + "right": "issue:47050", + "accept": false, + "reason": "Chunked generation with compiled forward and fp16 ONNX attention dtype mismatch are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "SigLIP2 documentation errors are not the same as the HF-vs-JAX implementation discrepancy." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Mamba2 use_cache correctness is unrelated to a T5 decoder input validation error message." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "PEFT local-checkpoint loading and offloading with quantization are different failures, even if both involve loading." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "A broken SigLIP2 model output and a CamemBERT masked-LM regression are unrelated model-specific bugs." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "T5 memory leak and missing supports_gradient_checkpointing on a trust_remote_code model are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:45923", + "accept": false, + "reason": "AutoModelForCausalLM.from_config not unwrapping text_config and a missing gradient-checkpointing flag are different issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: the cluster is mostly heterogeneous, 44514/44545 is a clear duplicate pair, and 42886/44843 is a plausible soft duplicate based on the offline-cache failure theme. The soft-edge rejections are conservative and avoid overclaiming broader equivalence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 18, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43618", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43756", + "issue:43819", + "issue:43873", + "issue:43927", + "issue:44008", + "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:46294", + "issue:46665", + "issue:46693", + "issue:46829", + "issue:46832", + "issue:46858", + "issue:46932", + "issue:47020", + "issue:47050" + ], + "soft_pairs": [ + "issue:45081|issue:45356", + "issue:44857|issue:47050", + "issue:44995|issue:46858", + "issue:45356|issue:45907", + "issue:43421|issue:44186", + "issue:39692|issue:43873", + "issue:39692|issue:47050", + "issue:43720|issue:43819", + "issue:43295|issue:45834", + "issue:43653|issue:43720", + "issue:46829|issue:46832", + "issue:43550|issue:43927", + "issue:45987|issue:46829", + "issue:43540|issue:46294", + "issue:43550|issue:44361", + "issue:44625|issue:47050", + "issue:45381|issue:46693", + "issue:44062|issue:44488", + "issue:43596|issue:44220", + "issue:43531|issue:43643", + "issue:43611|issue:46665", + "issue:41720|issue:44991", + "issue:44220|issue:44661", + "issue:44186|issue:44898", + "issue:30064|issue:36331", + "issue:43653|issue:43819", + "issue:44008|issue:46858", + "issue:45987|issue:46932", + "issue:43335|issue:43756", + "issue:43618|issue:43697", + "issue:44488|issue:47020", + "issue:43550|issue:43819" + ], + "prepared_review_unit_hash": "05d94adf35a3d606fa491fe4d8f43a221ce8be2162aaeb04a44ace3351517686", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37683, + "estimated_input_tokens": 9421, + "estimated_eval_tokens": 19098 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab bag of unrelated bug reports; the soft similarity signals are broad subsystem overlaps (tokenizers, caching, vision/video, quantization) rather than true duplicates. I do not see a single coherent duplicate cluster, and I would reject all listed soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45081 is the best anchor because it is a concrete, code-path-specific tokenizer regression in `_patch_mistral_regex`, and it is the only item with a plausible neighboring report in the same narrow area.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45081 is the most actionable and narrowly scoped bug report in the set; it describes a specific failure mode with a clear fix surface, unlike the broader or more model-specific reports around it.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both touch Mistral-related tokenizer regex handling, but they describe different failures and model contexts; this is related area, not the same underlying bug." + }, + { + "left": "issue:44857", + "right": "issue:47050", + "accept": false, + "reason": "Different models and different failure modes: AMP/CUDA loss crash versus fp16 ONNX attention dtype mismatch." + }, + { + "left": "issue:44995", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache-like behavior, but one is a stale indexer cache on a second forward pass while the other is a sequence-length tensor issue in static cache prefill arithmetic." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regression versus a CI helper bug forwarding revision; no shared underlying code-path problem." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Both are tokenizer issues, but one is about runtime special-token/post-processor updates and the other is a crash in NER/batched padding logic." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "Doc-example/model mismatch and quantization failure are unrelated to an offloading behavior regression." + }, + { + "left": "issue:39692", + "right": "issue:47050", + "accept": false, + "reason": "Documentation/model mismatch is unrelated to an ONNX dtype mismatch in attention." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "BitNet accelerated loading bug and DAC latent reconstruction bug are different models and different code paths." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor/tokenizer regression versus long OCR index error in a different model family." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBird special-token decode bug and BitNet packed-weight loading bug are unrelated." + }, + { + "left": "issue:46829", + "right": "issue:46832", + "accept": false, + "reason": "Different models and outputs: VideoPrism hidden_states assignment versus RT-DETR feature-level mismatch." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to DiaConfig token ID loss after save/load." + }, + { + "left": "issue:45987", + "right": "issue:46829", + "accept": false, + "reason": "Static cache sequence-length return type bug is unrelated to VideoPrism output field assignment." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both involve Qwen video input handling, but they are different models and distinct failure conditions." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Bamba compile/SDPA issue and MLuke tokenizer AttributeError are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:47050", + "accept": false, + "reason": "Configuration propagation bug versus ONNX attention dtype mismatch." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Qwen video-position-id regression and long-context FA2 performance regression are different problems." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken special-argument error is unrelated to a tokenizer loading failure for a specific model." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Deepspeed zero3/BertModel index error and fbank feature extraction issue are different code paths." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and AutoConfig trust_remote_code field loss are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:46665", + "accept": false, + "reason": "Base-model-prefix loading breakage and inconsistent from_pretrained behavior in classes are not the same defect." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction bug and tokenizer mapping-name failure are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "LayoutLMv2 tokenizer crash and Perceiver interpolation failure are different model-specific bugs." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image processor void segmentation-map bug and CustomTrainer signature mismatch are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "Tokenizer special-token registration bug and DAC latent conversion bug are unrelated." + }, + { + "left": "issue:44008", + "right": "issue:46858", + "accept": false, + "reason": "Gemma3n variable-name collision in forward is unrelated to static-cache sequence-length arithmetic." + }, + { + "left": "issue:45987", + "right": "issue:46932", + "accept": false, + "reason": "StaticCache return-type bug and Gemma4 quantized embedder dtype casting bug are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction bug and Smollm3 RoPE-layer counting bug are different configuration issues." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "CLIP attentions field regression and RT-DETR output differences are distinct model behaviors." + }, + { + "left": "issue:44488", + "right": "issue:47020", + "accept": false, + "reason": "Loading a specific tokenizer model and XLMRoberta tokenizer constructor type mismatch are different issues." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Bamba compile/SDPA problem and DAC latent mismatch are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 19, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:34634", + "issue:36010", + "issue:39692", + "issue:43322", + "issue:43388", + "issue:43421", + "issue:43540", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43643", + "issue:43644", + "issue:43749", + "issue:43782", + "issue:43818", + "issue:43825", + "issue:43994", + "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:46032", + "issue:46291", + "issue:46294", + "issue:46463", + "issue:46512", + "issue:47020" + ], + "soft_pairs": [ + "issue:43643|issue:44877", + "issue:44843|issue:45081", + "issue:44821|issue:45084", + "issue:39692|issue:45834", + "issue:43540|issue:44008", + "issue:34634|issue:36010", + "issue:44448|issue:44671", + "issue:44186|issue:44442", + "issue:45584|issue:45834", + "issue:45183|issue:45479", + "issue:43322|issue:45685", + "issue:44991|issue:47020", + "issue:43644|issue:43749", + "issue:43606|issue:45072", + "issue:43782|issue:44977", + "issue:43421|issue:44811", + "issue:44561|issue:45020", + "issue:44991|issue:46291", + "issue:44479|issue:45834", + "issue:43577|issue:43818", + "issue:43550|issue:45072", + "issue:43582|issue:46512", + "issue:43388|issue:46032", + "issue:43825|issue:43994", + "issue:46294|issue:46463", + "issue:44805|issue:44936" + ], + "prepared_review_unit_hash": "2b6bf7e879e2a15e1e34a14947da6d5862adb1308748cde14dca78ecb77d139f", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35859, + "estimated_input_tokens": 8965, + "estimated_eval_tokens": 18186 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose collection of mostly unrelated Transformers bug reports spanning tokenizers, multimodal models, dtype/loading regressions, and validation/error-message issues. I did not find any pair that clearly shares the same concrete underlying bug, so the soft edges should remain unmerged.", + "confidence": 0.71, + "canonical_issue_reason": "Issue 45020 is the broadest and most cluster-like anchor: it describes a general class of recent-version regressions affecting remote_code models, which is the closest thing here to an umbrella report.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45020 is the best single issue to keep as the representative cluster member because it is broad, well-scoped around a recurring regression theme, and most likely to subsume nearby remote_code-related reports without being too narrow.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both involve config loading, but one is about trust_remote_code returning incomplete configs and the other about strict config blocking granite_speech; different failure modes and code paths." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Both touch _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer attribute crash; not the same bug." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "Unrelated failures: AutoImageProcessor URL loading versus template-node compilation TypeError." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "Different multimodal model bugs on different models and paths; only loosely similar on the surface." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Completely different model families and error causes (video processing ValueError vs variable-name collision in Gemma3n forward)." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Both are v5 regression reports, but for different models and different inference behaviors; not a single fixable code-path problem." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "Tokenizer crash on NER/batched padding is unrelated to AutoTokenizer failing to load a specific tokenizer class." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Empty Whisper transcription after align_special_tokens is a different bug from Kosmos2.5 long OCR input indexing." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "One is input validation for input_ids/inputs_embeds, the other is loss computation for num_labels=1 classification; different subsystems." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Segfault in Llava Next loading and an MPS histogram dtype error are unrelated." + }, + { + "left": "issue:44991", + "right": "issue:47020", + "accept": false, + "reason": "Both are tokenizer load failures, but for different tokenizers and different root causes; too broad to merge." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer junk filling and FSDP CPU RAM efficient loading are distinct loading/runtime issues." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "Device mismatch under CPU offload is not the same as bfloat16 dtype mismatches in inference." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "from_pretrained weight_only loading error for Qwen3VL is unrelated to flash-attention generation behavior in Qwen3.5." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Special-token post-processor update bug and Whisper batch_decode skip_special_tokens behavior are different APIs and different bugs." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Related in theme, but 44561 is a specific breakage from removing is_torch_fx_available while 45020 is a broad umbrella report; not the same concrete issue." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading versus pipelines import failure due to kernels version/revision requirements; different failure paths." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Video-input regression across Qwen variants and long OCR index error in Kosmos2.5 are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "BLIP2 dtype inheritance and Video-LLaVA temporal-attention/weight-sharing are separate model bugs." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "torch.compile + SDPA failure in Bamba and bfloat16 dtype mismatch reports are not the same code path." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon TypeError in caching_allocator_warmup and stale model.config.dtype after from_config are distinct dtype-related issues." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Metric gathering drops labels in the last batch versus Mamba2Mixer use_cache correctness are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "A misleading pipeline error message and a nonsensical SigLIP2 output bug are different pipeline issues with different root causes." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen3.5 video chat-template token/feature mismatch and T5 decoder-input validation messaging are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask shape IndexError and trainer.evaluate() failing after train() do not share the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 20, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:42175", + "issue:43262", + "issue:43540", + "issue:43550", + "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:43906", + "issue:43931", + "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:45372", + "issue:45381", + "issue:45701", + "issue:46175", + "issue:46209", + "issue:46285", + "issue:46326", + "issue:46665", + "issue:46697", + "issue:46728", + "issue:46846", + "issue:46932" + ], + "soft_pairs": [ + "issue:43716|issue:44008", + "issue:43931|issue:46846", + "issue:43540|issue:45381", + "issue:28218|issue:29159", + "issue:44610|issue:45072", + "issue:43749|issue:44291", + "issue:43720|issue:45072", + "issue:43645|issue:46665", + "issue:43825|issue:43827", + "issue:44671|issue:45701", + "issue:43577|issue:43582", + "issue:34634|issue:36331", + "issue:42175|issue:45081", + "issue:46175|issue:46932", + "issue:46697|issue:46728", + "issue:45003|issue:46285", + "issue:43782|issue:45325", + "issue:43262|issue:43819", + "issue:43653|issue:44912", + "issue:43582|issue:44857", + "issue:45372|issue:46326", + "issue:44589|issue:45042", + "issue:43742|issue:45020", + "issue:43550|issue:44912", + "issue:43906|issue:46209", + "issue:36010|issue:43645" + ], + "prepared_review_unit_hash": "3746c10a6634784cc485ae21840a5883bc350cbe0c6935fd76290630e583883c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35824, + "estimated_input_tokens": 8956, + "estimated_eval_tokens": 18168 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a duplicate set: the soft pairs are mostly superficial matches across unrelated models, loaders, tokenizers, and processors. I would reject every soft edge and keep the issues separate.", + "confidence": 0.95, + "canonical_issue_reason": "No safe canonical issue: the items span many different code paths and models, and none of the soft pairs is a clear same-bug duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45701 is the broadest, highest-level regression description in the list, so it is the closest thing to a generic representative issue, but it is not a duplicate anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Both are multimodal model bugs, but one is an image preprocessor/model dtype mismatch and the other is a Gemma3n variable-collision AttributeError; different code paths and fixes." + }, + { + "left": "issue:43931", + "right": "issue:46846", + "accept": false, + "reason": "Both involve Qwen VL models, but one is a load-time weight-shape mismatch and the other is a tensor-parallel planning/OOM issue for a different model/version." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both mention video inputs in Qwen models, but the failures are different: a Qwen3OmniMoe ValueError vs a Qwen2.5-VL rope/indexing bug." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Both are tokenizer-related, but one is a specific added-token spacing bug while the other is a broad slow/fast tokenizer inconsistency; not clearly the same defect." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "These are unrelated: OmDet-Turbo image sizing versus dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Both concern model loading, but one is FSDP CPU RAM-efficient loading and the other is init_empty_weights / _is_hf_initialized handling; different underlying issues." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "Different problems: packed-weight unpacking during accelerate loading versus bfloat16 dtype mismatches." + }, + { + "left": "issue:43645", + "right": "issue:46665", + "accept": false, + "reason": "Both are about model initialization behavior, but one is a Jupyter/custom-model breakage and the other is from_pretrained inside Python classes; not the same bug." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "One is a runtime error message about unsupported translation tasks; the other is docs still referencing pipeline() after removals. Related theme, but not the same change." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM output regression and a generic tokenization-version change are not the same concrete failure." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "These are unrelated: Blip2 dtype propagation versus an Apple Silicon caching_allocator_warmup TypeError." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "No overlap in bug surface: BarkProcessor voice_preset versus CustomTrainer.compute_loss signature mismatch." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "These touch different subsystems entirely: backend installation extras versus a Mistral tokenizer regex/backend_tokenizer crash." + }, + { + "left": "issue:46175", + "right": "issue:46932", + "accept": false, + "reason": "Both are Gemma4 embedder issues, but one is a huge one-hot position-embedding materialization problem and the other is dtype casting for quantized embedders." + }, + { + "left": "issue:46697", + "right": "issue:46728", + "accept": false, + "reason": "Both are image preprocessing bugs, but they affect different models and different calculations (padding vs patch counting)." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to CLIPTextModel structural inconsistency." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen VL bugs: one is from_pretrained weight_only loading, the other is rope/index computation for still-image temporal positions." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate defaults and DAC.from_latents STE mismatch are different audio-processing problems." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Apple Silicon allocator warmup and CUDA float16 AMP loss crashes are unrelated failures." + }, + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "Both involve mistral_common integration, but one is an import failure for ReasoningEffort and the other is continue_final_message behavior; distinct code paths." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup failure and PIL backend torchvision requirement are completely different bugs." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "A key error loading MobileLLM-125M is not the same as the broad remote_code compatibility regression." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "An isolated reproduction of an earlier issue is not the same as an FP8 inference failure on B200." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import failure and Jupyter custom-model initialization breakage are different issues, even if both relate to v5 API changes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 21, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:34567", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43450", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43638", + "issue:43643", + "issue:43696", + "issue:43697", + "issue:43742", + "issue:43746", + "issue:43819", + "issue:43854", + "issue:43927", + "issue:43931", + "issue:44060", + "issue:44368", + "issue:44387", + "issue:44479", + "issue:44554", + "issue:44610", + "issue:44617", + "issue:44811", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45081", + "issue:45084", + "issue:45406", + "issue:45538", + "issue:45600", + "issue:45759", + "issue:45859", + "issue:46291", + "issue:47050" + ], + "soft_pairs": [ + "issue:43526|issue:43697", + "issue:43257|issue:43329", + "issue:36010|issue:36331", + "issue:43452|issue:44991", + "issue:45081|issue:45406", + "issue:43638|issue:44936", + "issue:43531|issue:45042", + "issue:44060|issue:44368", + "issue:42175|issue:46291", + "issue:43819|issue:43927", + "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:45859|issue:47050", + "issue:43643|issue:45759", + "issue:38175|issue:44610", + "issue:39692|issue:44912", + "issue:43257|issue:43696", + "issue:43262|issue:44811", + "issue:44387|issue:44617", + "issue:43257|issue:43931", + "issue:43329|issue:43450", + "issue:43329|issue:44479", + "issue:42915|issue:43742", + "issue:34567|issue:40990" + ], + "prepared_review_unit_hash": "529a20d51424c14a3b788c9a8524d043f00c5d3c1eff283dc3a94a4103e32bf7", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36454, + "estimated_input_tokens": 9114, + "estimated_eval_tokens": 18484 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous and does not contain clear duplicate pairs; the soft-similarity matches are mostly same-subsystem, not same-bug, relationships.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43257 is the strongest representative issue here: it describes a concrete, reproducible Qwen3-MoE loading failure with accelerate+deepspeed, has substantial discussion, and is more actionable than the other heterogeneous reports.", + "canonical_pr_reason": "No PRs are present in this cluster.", + "best_issue_reason": "issue:43257 is the best overall issue candidate because it is specific, technically grounded, and appears to capture a real code-path bug rather than a vague symptom or documentation problem.", + "best_pr_reason": "No PRs are present in this cluster.", + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "Different bugs: BeitImageProcessorFast label reduction vs RTDetrV2 output differences under v5." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Both involve Qwen/MoE-related code, but one is a loading/conversion bug and the other is a multimodal token-counting video-branch bug." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "Unrelated failures: missing GenerationMixin import vs CustomTrainer.compute_loss signature mismatch." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/loading issues, but they point to different failure modes and different affected code paths." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Different components: Mistral regex patching in tokenizer internals vs Gemma4Processor missing _tokenizer in serve." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Distinct training/eval failures with different triggers and symptoms." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior is unrelated to PIL image processors requiring torchvision." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both are warning/tie-weights related, but they concern different models and different root causes." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Packaging/backend selection bug vs pipelines import failure due to kernels version/revision requirements." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "Different model/config bugs: DAC latent reconstruction mismatch vs DiaConfig token-ID persistence loss." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template compilation error and tokenizer max_length misuse are unrelated." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Video multimodal token counting bug vs stale auto_mappings references to removed configs." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Separate issues: Beit label reduction vs MPS attention correctness when head dims differ." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Local-path import warning is unrelated to Qwen3-MoE weight conversion during accelerate+deepspeed loading." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT/local-checkpoint loading bug vs Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "GLM-4.7-Flash test loading issue vs Qwen3.5 flash-attention generation regression." + }, + { + "left": "issue:45859", + "right": "issue:47050", + "accept": false, + "reason": "CPU offload incompatibility in Qwen3_5Moe rotary embedding is unrelated to ONNX export dtype mismatch in CanineModel." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss is a different bug from unwrap text_config in composite multimodal configs." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "SigLIP2 zero-probability behavior and OmDet-Turbo processor/model size mismatch are different problems." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Documentation-example model/processor issues are unrelated to MXFP4 quantization fallback on gpt-oss-20b." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3-MoE weight conversion failure is not the same as GPT-oss-20b OOM." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat-template sampling-rate default bug is unrelated to Whisper decode token filtering." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both mention OOM, but one is an int4 quantization reserved-memory regression and the other is a Sam3Video memory limit issue." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Different Qwen model families and different failures: accelerate/deepspeed loading vs weight-shape mismatch." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Both are video-related, but one is an internal token-counting undefined-symbol bug and the other is a batched output-shape bug in video processors." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Related domain, but not the same bug: undefined helper usage in token counting vs broader v5.3.0 video-input regression across several models." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure in Qwen3Moe is unrelated to a MobileLLM-125M key error." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token counter not updating is unrelated to high perplexity on GPT-oss-20b." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 22, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:29159", + "issue:29870", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:42175", + "issue:42915", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43525", + "issue:43540", + "issue:43577", + "issue:43650", + "issue:43697", + "issue:43749", + "issue:43824", + "issue:43906", + "issue:43927", + "issue:44291", + "issue:44479", + "issue:44568", + "issue:44756", + "issue:44821", + "issue:44857", + "issue:44912", + "issue:44987", + "issue:45042", + "issue:45081", + "issue:45127", + "issue:45137", + "issue:45356", + "issue:45357", + "issue:45405", + "issue:45459", + "issue:45538", + "issue:45803", + "issue:46291", + "issue:46421", + "issue:46424", + "issue:46693", + "issue:46701", + "issue:46832", + "issue:46869", + "issue:47020" + ], + "soft_pairs": [ + "issue:44291|issue:45042", + "issue:46421|issue:46424", + "issue:45459|issue:45803", + "issue:42915|issue:44821", + "issue:43650|issue:43906", + "issue:29159|issue:34567", + "issue:43749|issue:44756", + "issue:43540|issue:46701", + "issue:43525|issue:46291", + "issue:28282|issue:29870", + "issue:44568|issue:45356", + "issue:44987|issue:45538", + "issue:43454|issue:43577", + "issue:43927|issue:44912", + "issue:28282|issue:34689", + "issue:43425|issue:46291", + "issue:33357|issue:34567", + "issue:43697|issue:45357", + "issue:44857|issue:46832", + "issue:43404|issue:45127", + "issue:42175|issue:43824", + "issue:45081|issue:47020", + "issue:45137|issue:45405", + "issue:28282|issue:29127", + "issue:42915|issue:45538", + "issue:44479|issue:46693", + "issue:45538|issue:46869" + ], + "prepared_review_unit_hash": "d904bdb6356e0b64e8efc00cefb63df8015c8ce4cb6ffae1c74ada682e065872", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35426, + "estimated_input_tokens": 8857, + "estimated_eval_tokens": 17970 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is highly heterogeneous: it spans unrelated tokenizer, loading, quantization, training, vision, and performance bugs. None of the suggested soft pairs look like true duplicates, so all soft edges are rejected.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 45042 is the clearest, well-scoped, user-facing regression report among an otherwise unrelated set, but there is no strong duplicate cluster here.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45042 is the best anchor only in the sense that it is concrete, reproducible, and clearly describes an actionable bug; it is not actually a duplicate hub for the other items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Different failures: `init_empty_weights` argument handling vs incorrect torchvision dependency for PIL image processors." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache, but one is lazy-init recompilation and the other is `max_cache_length` being honored only for quantized cache." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Different code paths: tokenizer protobuf error masking vs a bare `except:` in Fuyu tensor conversion." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Unrelated bugs: FineGrainedFP8Config failure for Qwen3Moe vs loading `AutoImageProcessor` from URL." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "No overlap in underlying bug; one is a vague placeholder issue, the other is a reproduction for a prior bug." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency is unrelated to `TrainerState.num_input_tokens_seen` not updating." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "Different symptoms and mechanisms: FSDP CPU RAM efficient loading vs disabling mmap on Strix Halo to avoid OOM." + }, + { + "left": "issue:43540", + "right": "issue:46701", + "accept": false, + "reason": "Both involve inputs/pipelines, but one is Qwen3OmniMoe video input validation and the other is a DiffusionGemma pipeline output issue." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Llama4 `pad_token_id` attribute missing is unrelated to kernels v0.15 breaking `transformers.pipelines` import." + }, + { + "left": "issue:28282", + "right": "issue:29870", + "accept": false, + "reason": "Completely different errors: missing PyTorch ImportError vs BART `decoder_attention_mask` unbound local in `generate()`." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but one is special-token insertion and the other is Kimi-K2.5 codec/fix_mistral_regex behavior." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Different tokenizer/model issues: loading `physical-intelligence/fast` vs CLIPTokenizer's `model_max_length` sentinel value." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "AyaVision weight tying bug is unrelated to BLIP2 dtype staying float32." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Different bugs: DiaConfig save/load token IDs vs MXFP4 quantization falling back to bf16." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Missing PyTorch ImportError is unrelated to Llama 3.2 Vision model loading failure." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Dependency/version problems, but not the same one: Torch 2.10 incompatibility vs kernels v0.15 import failure." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS CLIP bus error is unrelated to `num_input_tokens_seen` not updating in TrainerState." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different underlying bugs: RT-DETR output differences in v5 vs Qwen3.5 save_pretrained visual encoder keys." + }, + { + "left": "issue:44857", + "right": "issue:46832", + "accept": false, + "reason": "Both are RT-DETR/Lw-DETR related, but one is AMP/CUDA loss crashing and the other is a forward shape/feature-level mismatch." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both involve tied weights, but one is missing `lm_head` tying in Mistral3 and the other is LoRA merge with extended vocabulary causing collapse." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Different issues: missing TensorFlow in a torch install extra vs failing to import `Qwen2_5_VLForConditionalGeneration`." + }, + { + "left": "issue:45081", + "right": "issue:47020", + "accept": false, + "reason": "Tokenizer internals differ: Mistral regex/backend tokenizer crash vs XLMRoberta passing a dict to Unigram." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT minimum-version bump." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Different problems: missing PyTorch dependency vs unclear error message for LayoutLMv3 box content." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to CLIPTokenizer's `model_max_length` value." + }, + { + "left": "issue:44479", + "right": "issue:46693", + "accept": false, + "reason": "One is a video-input regression across Qwen models; the other is a long-context FA2 performance regression." + }, + { + "left": "issue:45538", + "right": "issue:46869", + "accept": false, + "reason": "CLIPTokenizer configuration issue is unrelated to YOLO multi-GPU device-map loading 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": 26, + "review_unit_index": 23, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:24643", + "issue:27957", + "issue:28282", + "issue:29159", + "issue:34567", + "issue:36331", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43452", + "issue:43531", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43716", + "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:45325", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45600", + "issue:45701", + "issue:45907", + "issue:46399", + "issue:46512", + "issue:46701", + "issue:46821" + ], + "soft_pairs": [ + "issue:44936|issue:46821", + "issue:43742|issue:44291", + "issue:44062|issue:44991", + "issue:24643|issue:29159", + "issue:44849|issue:45701", + "issue:43452|issue:43582", + "issue:42915|issue:43749", + "issue:43577|issue:46512", + "issue:27957|issue:28282", + "issue:41720|issue:43540", + "issue:43540|issue:44560", + "issue:44987|issue:45907", + "issue:42175|issue:44488", + "issue:44479|issue:45381", + "issue:43540|issue:44479", + "issue:38175|issue:44898", + "issue:34567|issue:36331", + "issue:44568|issue:45488", + "issue:44291|issue:44821", + "issue:44792|issue:44964", + "issue:43257|issue:43756", + "issue:44062|issue:45081", + "issue:45042|issue:45081", + "issue:43582|issue:46821", + "issue:43742|issue:44589", + "issue:43756|issue:46399", + "issue:43756|issue:45325", + "issue:43531|issue:44821", + "issue:43716|issue:46701", + "issue:45405|issue:45600" + ], + "prepared_review_unit_hash": "ccbba8c4882103e9064c5e590de88945de8f376985137f5ed097f7b2a84d269c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36681, + "estimated_input_tokens": 9171, + "estimated_eval_tokens": 18598 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are clearly different bugs touching different models/code paths. The only strong duplicate match is the Qwen2.5-VL/Qwen3-VL video-input regression pair, where one issue is a broader family report and the other is a specific manifestation of the same position-id bug.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 44479 is the best canonical issue because it is the broader regression report for the v5.3.0 Qwen video-input bug family and subsumes the more specific Qwen2.5-VL position-id report.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44479 is the strongest issue anchor in this set: it names the broader affected family, has clear regression framing, and matches the only accepted duplicate edge.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44936", + "right": "issue:46821", + "accept": false, + "reason": "Different failures in different areas: trainer evaluate state vs TimesFM window_size handling." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Both are load-time TypeErrors, but they involve different mechanisms and different models/code paths." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading problems, but the symptoms and affected models are unrelated." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed weight shape error and tokenizer slow/fast inconsistency are unrelated bugs." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "One is a Qwen3.5 hidden-states bug; the other is a broad tokenization-version complaint." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "AutoTokenizer/AutoModel gguf handling and Apple Silicon allocator warmup TypeError are different subsystems." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "Different training/loading failures: FP8 config handling vs FSDP CPU RAM efficient loading." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Both concern dtype state, but one is model/submodule dtype loading and the other is config.dtype staleness in from_config." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention support and missing PyTorch detection are unrelated issues." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Different Qwen model families and different failure modes: device-mapping CUDA assert vs video input ValueError." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both mention video inputs, but they affect different Qwen variants with different errors and likely different fixes." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "A model load failure and list_repo_files not forwarding revision are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Package/backend dependency packaging and a model load failure are not the same bug." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Both report the same Qwen video-input regression in transformers 5.3.0; 45381 is a specific symptom of the broader bug in 44479." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Different vision-model failures: unexpected zero probabilities vs Perceiver resize/interpolation failure." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token counting and custom compute_loss signature mismatch are separate trainer issues." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but they involve different tokenizers and distinct behavior changes." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Init-empty-weights argument handling and loading AutoImageProcessor from URL are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal regressions with different entry points: a Janus test failure vs Phi-4 multimodal load failure." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MoE weight conversion and Smollm3 RoPE-layer dropping are unrelated model-implementation bugs." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer bugs: AddedToken special-arg duplication vs Mistral regex patch failure." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "PIL image-processor dependency requirements and Mistral tokenizer regex patching are unrelated." + }, + { + "left": "issue:43582", + "right": "issue:46821", + "accept": false, + "reason": "Allocator warmup on Apple Silicon and TimesFM window_size AttributeError are different problems." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Different loading failures: MobileLLM key error vs missing Float8 storage object." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "RoPE-layer dropping in Smollm3 and a Qwen3.5 kernel registration crash are unrelated." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Both concern rotary/position handling, but they affect different models and different bugs." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and AutoImageProcessor URL loading are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:46701", + "accept": false, + "reason": "Mistral-3 dtype mismatch and DiffusionGemma pipeline failure are different code paths." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "PEFT version bump and removed Sam3LiteText auto-mapping references are different maintenance issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on the soft edges. The one accepted duplicate pair, issue:44479 and issue:45381, is well supported by the titles and the stated v5.3.0 Qwen video-input regression family. The remaining pair rejections are cautious and based on clear differences in model family, subsystem, or failure mode rather than overclaiming equivalence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 24, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29870", + "issue:32090", + "issue:41762", + "issue:43452", + "issue:43526", + "issue:43540", + "issue:43577", + "issue:43643", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43749", + "issue:43906", + "issue:44466", + "issue:44488", + "issue:44493", + "issue:44568", + "issue:44625", + "issue:44737", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44936", + "issue:45005", + "issue:45072", + "issue:45200", + "issue:45216", + "issue:45357", + "issue:45375", + "issue:45479", + "issue:45588", + "issue:45663", + "issue:45672", + "issue:45674", + "issue:45823", + "issue:45920", + "issue:46050", + "issue:46121", + "issue:46463", + "issue:46468", + "issue:46491", + "issue:46693", + "issue:46829", + "issue:46846", + "issue:46899" + ], + "soft_pairs": [ + "issue:44737|issue:46463", + "issue:44568|issue:46491", + "issue:45823|issue:46463", + "issue:43906|issue:46693", + "issue:43577|issue:45072", + "issue:45672|issue:45674", + "issue:24643|issue:29870", + "issue:43906|issue:46468", + "issue:43540|issue:46693", + "issue:43452|issue:44488", + "issue:44625|issue:44877", + "issue:45663|issue:46846", + "issue:43697|issue:46829", + "issue:46463|issue:46899", + "issue:28282|issue:32090", + "issue:44913|issue:46121", + "issue:44737|issue:45588", + "issue:43526|issue:45479", + "issue:45357|issue:45920", + "issue:43697|issue:45216", + "issue:45200|issue:46463", + "issue:43643|issue:45375", + "issue:43720|issue:44898", + "issue:44466|issue:45005", + "issue:41762|issue:44936", + "issue:44493|issue:46050", + "issue:43688|issue:43749" + ], + "prepared_review_unit_hash": "37d41222822a9848ef4dbebcee3a9587bc2d720ef470d9db1cb437f8c4bf4b82", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35916, + "estimated_input_tokens": 8979, + "estimated_eval_tokens": 18214 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated one-off bugs across different models, loaders, and training paths. I did not find any soft pair that clearly describes the same underlying bug/change, though a few are loosely similar in theme (e.g. tied-weights or config regressions).", + "confidence": 0.78, + "canonical_issue_reason": "Issue #45005 is the broadest and most representative report in the set, covering a concrete v5 tied-weights regression with detailed reproduction context; it is the closest thing to a central theme here.", + "canonical_pr_reason": null, + "best_issue_reason": "#45005 is the best single issue to anchor this cluster because it is substantial, specific, and captures the most plausible overlap with another tied-weights report without being overly narrow.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "Different models and different failures: XLNet CPU device placement vs T5 decoder input validation messaging." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and symptoms, so they are not the same bug." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 device-map forward mismatch is unrelated to T5 decoder input validation." + }, + { + "left": "issue:43906", + "right": "issue:46693", + "accept": false, + "reason": "A reproduction of an older issue is not the same as a FA2 long-context performance regression." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype issues, but they concern different models and different code paths." + }, + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Both are 4-bit inference issues, but one is a uint8/torch.finfo crash in Gemma4 audio modules and the other is a BitsAndBytes skip-module exclusion bug." + }, + { + "left": "issue:24643", + "right": "issue:29870", + "accept": false, + "reason": "DeepSpeed training weight-shape error and BART generate() unbound variable are unrelated failures." + }, + { + "left": "issue:43906", + "right": "issue:46468", + "accept": false, + "reason": "An isolated reproduction of one bug is not the same as a broken Nemotron example." + }, + { + "left": "issue:43540", + "right": "issue:46693", + "accept": false, + "reason": "Video input processing failure in Qwen3OmniMoe is unrelated to FA2 performance regression." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but one is a gguf_file API break and the other is a specific model load failure; no clear same code-path." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "One is num_labels propagation in Qwen3.5, the other is strict config handling for granite_speech." + }, + { + "left": "issue:45663", + "right": "issue:46846", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states bug and Qwen3.5 TP plan omissions are different distributed-training problems." + }, + { + "left": "issue:43697", + "right": "issue:46829", + "accept": false, + "reason": "Different models and outputs: RTDetrV2 v5 regression vs VideoPrism hidden_states field mix-up." + }, + { + "left": "issue:46463", + "right": "issue:46899", + "accept": false, + "reason": "T5 decoder input error messaging is unrelated to Gemma4 multimodal dtype casting under 4-bit quantization." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Missing PyTorch import and a Trainer broadcast type error are unrelated." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Both touch config/rope-related settings, but one is GPTNeoX reload behavior and the other is a utility type error on JSON-loaded lists." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU arange placement and flash-attention sink-less model crash are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and sequence-classification loss selection are unrelated bugs." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "One is Qwen3.5 save_pretrained visual-key serialization; the other is tokenizer token-ID generation across multiple models." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "RTDetrV2 output drift and Qwen3.5 checkpoint serialization regression are different issues." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Gemma 4 text-only mm_token_type_ids behavior is unrelated to T5 decoder input validation errors." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Both involve missing config fields, but one is generic trust_remote_code loading and the other is a specific Qwen3_5MoeVisionConfig strict-field issue." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific bugs." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both concern tied weights, but one is inconsistent lm_head serialization by device and the other is a broader translation-model regression; not enough evidence they are the same bug." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading failure and trainer.evaluate() after train() are unrelated." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Unexpected position-id keys and sinq save/load breakage are different serialization problems." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 25, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:29159", + "issue:33357", + "issue:38175", + "issue:42898", + "issue:43295", + "issue:43452", + "issue:43525", + "issue:43531", + "issue:43577", + "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:44479", + "issue:44568", + "issue:44625", + "issue:44743", + "issue:44811", + "issue:44821", + "issue:44843", + "issue:44857", + "issue:44898", + "issue:44991", + "issue:45356", + "issue:45372", + "issue:45542", + "issue:45701", + "issue:45704", + "issue:46170", + "issue:46178", + "issue:46402", + "issue:46491", + "issue:46701", + "issue:46821", + "issue:46829", + "issue:46832", + "issue:46932", + "issue:47000", + "issue:47013" + ], + "soft_pairs": [ + "issue:38175|issue:43295", + "issue:43452|issue:44843", + "issue:44361|issue:44442", + "issue:43653|issue:44361", + "issue:43761|issue:46829", + "issue:43819|issue:44811", + "issue:43994|issue:44898", + "issue:45372|issue:46170", + "issue:44857|issue:47000", + "issue:43531|issue:43742", + "issue:46832|issue:46932", + "issue:43525|issue:46701", + "issue:43295|issue:45356", + "issue:29159|issue:33357", + "issue:45542|issue:46178", + "issue:43653|issue:44568", + "issue:43688|issue:45704", + "issue:43688|issue:45701", + "issue:43646|issue:43749", + "issue:43756|issue:46402", + "issue:42898|issue:46491", + "issue:44464|issue:44743", + "issue:44843|issue:44991", + "issue:43761|issue:47013", + "issue:43452|issue:43577", + "issue:43653|issue:43927", + "issue:44479|issue:44625", + "issue:44821|issue:46821" + ], + "prepared_review_unit_hash": "7fa0c5464532ebcd888ccf091148c103e6e3e01f347c0e43748d8d3117b7bad0", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36305, + "estimated_input_tokens": 9077, + "estimated_eval_tokens": 18410 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is a heterogeneous mix of Transformers bugs, with a large tokenizer/regression subtheme but no pair that clearly shares the same concrete root cause. Under a strict duplicate standard, I would reject all soft edges.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45701 is the broadest, most representative report in the set: it describes version-dependent tokenization changes and sits near the center of the tokenizer-related issues, while many others are narrower model-specific regressions.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45701 is the best single issue to represent the cluster because it is generic, user-facing, and closest to the recurring tokenizer/regression theme that appears in several other items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Different failure modes: one is zero probabilities for SigLIP2, the other is processor/tokenizer regression with images. Same broad area, not the same bug." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both involve tokenizer loading, but one is gguf/from_pretrained breakage and the other is offline mode due to an unconditional hub call. Different root causes." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both are AttributeErrors in tokenizers, but they affect different tokenizers and loading paths. Not the same underlying issue." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "BigBirdTokenizer special-token registration/empty decode is unrelated to MLukeTokenizer task-time AttributeError." + }, + { + "left": "issue:43761", + "right": "issue:46829", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression is a vision-model output bug, while VideoPrism returns the wrong field in an output object. Different code paths." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latents/STE mismatch is unrelated to Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Both are model inference issues, but SigLIP2 nonsensical outputs and Perceiver interpolation failure have different causes and affected components." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Gemma processor loading import failure and Windows/ROCm distributed import error are distinct dependency problems." + }, + { + "left": "issue:44857", + "right": "issue:47000", + "accept": false, + "reason": "LwDetr float16 CUDA crash and RT-DETR cost-matrix infeasibility are separate loss/assignment failures." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM key error are unrelated configuration/loading bugs." + }, + { + "left": "issue:46832", + "right": "issue:46932", + "accept": false, + "reason": "RT-DETR backbone feature mismatch and Gemma4 quantized embedder dtype casting are different model-specific issues." + }, + { + "left": "issue:43525", + "right": "issue:46701", + "accept": false, + "reason": "Llama4Config pad_token_id missing and DiffusionGemma pipeline failure do not share a root cause." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Both mention processor/tokenizer regressions, but one is custom processor/tokenizer access and the other is Kimi-K2.5 codec/fix_mistral_regex behavior. Different defects." + }, + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Slow vs fast tokenizer inconsistency is a tokenizer API issue; bus error on MacOS with CLIP is a platform crash. No shared bug." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Undefined tf backend from tensorboard-only install and BACKENDS_MAPPING tensorflow_text error are separate backend-resolution problems." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and mdeberta-v3-base add_special_tokens behavior are both tokenizer issues but not the same tokenizer or fix." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Auxiliary loss normalization in MoE models is unrelated to T5 apex FusedRMSNorm memory leak." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Auxiliary loss normalization and tokenization-version drift are different subsystems and bugs." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Transformers 5.0 custom model initialization breakage and FSDP CPU RAM efficient loading are distinct upgrade regressions." + }, + { + "left": "issue:43756", + "right": "issue:46402", + "accept": false, + "reason": "Smollm3 RoPE-layer mismatch and Sam3 text encoder weights not loading are unrelated model definition/loading issues." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "clean_up_tokenization_spaces v5 behavior change and CodeLlama leading-space round-trip regression are both tokenizer-related but different concrete behaviors." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and Qwen3.5 recurrent state reset are separate generation/state-management bugs." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Offline-mode failure from an unconditional model_info call and EMBEDDIA tokenizer loading failure are not the same code-path problem." + }, + { + "left": "issue:43761", + "right": "issue:47013", + "accept": false, + "reason": "CLIPVision hidden_states regression and xLSTM train-mode unpacking crash are unrelated output/state handling bugs." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading breakage and Blip2 dtype propagation are different from-pretrained/model-initialization issues." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and DiaConfig custom token IDs not persisting are separate tokenizer/config persistence bugs." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video input regression for Qwen3.5-family models and num_labels propagation in Qwen3.5 are different model configuration issues." + }, + { + "left": "issue:44821", + "right": "issue:46821", + "accept": false, + "reason": "AutoImageProcessor URL loading and TimesFM window_size AttributeError are unrelated loader/config problems." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 26, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:32090", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43611", + "issue:43643", + "issue:43688", + "issue:43720", + "issue:43782", + "issue:43825", + "issue:43927", + "issue:44361", + "issue:44464", + "issue:44466", + "issue:44484", + "issue:44514", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44625", + "issue:44704", + "issue:44734", + "issue:44964", + "issue:44977", + "issue:45290", + "issue:45405", + "issue:45440", + "issue:45491", + "issue:45584", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46209", + "issue:46365", + "issue:46612", + "issue:46693", + "issue:46846" + ], + "soft_pairs": [ + "issue:33666|issue:36010", + "issue:44484|issue:44704", + "issue:44466|issue:46050", + "issue:32090|issue:33666", + "issue:43688|issue:45440", + "issue:44734|issue:46612", + "issue:44977|issue:46693", + "issue:41720|issue:44464", + "issue:44464|issue:44977", + "issue:43504|issue:43550", + "issue:43278|issue:43688", + "issue:43504|issue:43720", + "issue:43782|issue:46693", + "issue:45491|issue:46002", + "issue:43262|issue:43295", + "issue:33666|issue:36331", + "issue:43611|issue:46846", + "issue:43577|issue:44554", + "issue:43720|issue:43927", + "issue:44514|issue:45290", + "issue:43606|issue:43927", + "issue:43504|issue:44610", + "issue:43643|issue:44561", + "issue:43525|issue:43572", + "issue:44964|issue:45405", + "issue:43720|issue:44361", + "issue:44625|issue:45584", + "issue:44977|issue:46209", + "issue:43493|issue:43825", + "issue:44554|issue:46365", + "issue:45823|issue:46050" + ], + "prepared_review_unit_hash": "38cd7d2389c3ef72371ac4f9ea52d9c67e850930c68dc0b65238c25dd0d202c7", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37142, + "estimated_input_tokens": 9286, + "estimated_eval_tokens": 18828 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly distinct issues with a few thematic overlaps (v5 regressions, chat-template crashes, save/load bugs), but none of the soft pairs look like the same concrete bug/change.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43611 is the clearest broad framework regression in the set: it names a concrete v5 loading breakage with a shared root cause and wider applicability than the model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best representative issue overall because it is concrete, broadly impactful, and easy to triage against related v5 compatibility regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Different failures: multi-GPU Qwen2-VL training vs a GenerationMixin import error." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "Unrelated topics: shard-size question vs AutoProcessor kwargs forwarding." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both involve save/load, but one is tied-weight serialization and the other is quantized model persistence." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is not the same as Qwen2-VL multi-GPU training support." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both are MoE-related, but one is auxiliary-loss normalization and the other is a native-vs-remote implementation mismatch." + }, + { + "left": "issue:44734", + "right": "issue:46612", + "accept": false, + "reason": "Different generation bugs: KV-cache continuation indexing vs beam-search cache reorder omission." + }, + { + "left": "issue:44977", + "right": "issue:46693", + "accept": false, + "reason": "Flash-attention correctness failure is distinct from a long-context FA2 performance regression." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Auto device-map cuda assert is unrelated to chunked generation with compiled forward." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Different symptoms and code paths: compiled-forward inconsistency vs flash-attention generation failure." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Legacy-field loading bug for BEiT is unrelated to Bamba torch.compile/SDPA failure." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Embedding dtype drift during eval is not the same bug as MoE auxiliary-loss normalization." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Pretrained preset loading failure and packed-weight accelerate loading failure are different issues." + }, + { + "left": "issue:43782", + "right": "issue:46693", + "accept": false, + "reason": "Qwen3VL weight_only load error is different from FA2 long-context performance regression." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both mention NaNs, but one is sliding-window attention on GPU and the other is classification-head init/checkout." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Audio processor sampling-rate default bug is unrelated to processor.tokenizer regression." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Multi-GPU training request is not the same as a CustomTrainer compute_loss signature break." + }, + { + "left": "issue:43611", + "right": "issue:46846", + "accept": false, + "reason": "Both are v5/model-loading related, but base_model_prefix loading and Qwen3.5 TP-plan linear_attn are separate bugs." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Blip2 dtype propagation and MPS attention correctness are different failure modes." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Packed-weight accelerate loading and DiaConfig token-ID persistence are unrelated serialization bugs." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both hit apply_chat_template, but the triggering conditions and failure paths differ." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to DiaConfig save/load token-ID loss." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Legacy-field model loading and OmDet-Turbo processor size mismatch are different issues." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both affect trust_remote_code, but one is missing config fields and the other is removal of a utility import." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Similar theme, but different model configs and different missing fields; not one underlying bug." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model-loading failure and unreleased PEFT-version bump are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Different areas: accelerate weight unpacking vs tokenizer AttributeError." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Qwen3.5 num_labels propagation is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug is not the same as FP8 inference failure on B200." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message wording are unrelated." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "MPS attention correctness and Trainer import failure due to datasets are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Gemma4 device-map forward mismatch is not the same as quantized model save/load breakage." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 27, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28218", + "issue:33357", + "issue:33666", + "issue:39692", + "issue:42796", + "issue:43257", + "issue:43278", + "issue:43295", + "issue:43299", + "issue:43334", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43653", + "issue:43701", + "issue:43824", + "issue:43844", + "issue:43927", + "issue:43994", + "issue:44291", + "issue:44361", + "issue:44451", + "issue:44466", + "issue:44488", + "issue:44493", + "issue:44568", + "issue:44589", + "issue:44792", + "issue:44843", + "issue:44849", + "issue:44977", + "issue:45245", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45759", + "issue:45812", + "issue:46192", + "issue:46209", + "issue:46285", + "issue:46489", + "issue:46710", + "issue:46821", + "issue:46832" + ], + "soft_pairs": [ + "issue:28218|issue:33357", + "issue:43452|issue:44451", + "issue:43994|issue:46192", + "issue:42796|issue:44977", + "issue:39692|issue:43295", + "issue:43493|issue:43844", + "issue:43278|issue:43844", + "issue:46209|issue:46710", + "issue:44849|issue:46209", + "issue:43646|issue:44466", + "issue:44568|issue:46489", + "issue:43278|issue:43701", + "issue:44568|issue:45812", + "issue:33357|issue:33666", + "issue:44488|issue:44843", + "issue:44792|issue:45405", + "issue:43927|issue:44361", + "issue:44849|issue:45357", + "issue:43824|issue:44589", + "issue:43575|issue:45405", + "issue:46821|issue:46832", + "issue:45245|issue:46710", + "issue:43572|issue:44291", + "issue:43334|issue:45759", + "issue:43504|issue:43606", + "issue:44589|issue:45520", + "issue:43334|issue:43572", + "issue:43257|issue:43299", + "issue:44493|issue:46285", + "issue:43504|issue:43653" + ], + "prepared_review_unit_hash": "02e888d06bdb2d5e9e494351a1c86b6f1249e92236486e1448218429d3b9e053", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36856, + "estimated_input_tokens": 9214, + "estimated_eval_tokens": 18684 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated regressions across different models/subsystems; despite some tokenizer/loading overlap, none of the soft pairs look like the same underlying bug/change. No PRs are present in the cluster.", + "confidence": 0.97, + "canonical_issue_reason": "issue:45812 is the strongest representative: it describes a concrete, high-impact tokenizer regression with broad scope (all Granite models), a clear version boundary, and a silent behavioral change.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is a good global anchor because it is specific, user-facing, and clearly reproducible, making it more representative than narrow model-specific or environment-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer spacing bug vs MacOS bus error in CLIP loading; different symptoms and code paths." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "Both are loading failures, but one is gguf/AutoTokenizer/AutoModel behavior and the other is a specific pretrained model compatibility issue." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 inference quality issue vs return_tensors being ignored in multimodal input handling; not the same bug." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer BPE marker handling vs Qwen3.5 flash-attention generation bug; unrelated failures." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Documentation example errors vs a regression in processor/tokenizer access; different concrete problems." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 JAX parity discrepancy vs a DeepSpeed/ZeRO gradient instability; different layers of the stack." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Training/eval dtype mismatch vs gradient explosion under ZeRO-3; not the same underlying defect." + }, + { + "left": "issue:46209", + "right": "issue:46710", + "accept": false, + "reason": "FP8 inference failure on B200 vs accuracy regression after version upgrade; different hardware/pathology." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen hidden-states bug vs FP8 inference issue; no shared code-path evidence." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model initialization breakage vs serialization inconsistency of tied weights; distinct regression types." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "mDeBERTa special-token addition bug vs DeepSeek-Coder tokenizer output regression; different tokenizers and causes." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "Embedding dtype mismatch vs resume checkpoint key mismatch; unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer-related, but one concerns special-token addition behavior and the other wrong token IDs for Granite models." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training support request; unrelated." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Model load failure for a specific checkpoint vs offline-mode breakage from unconditional model_info() lookup; different root causes." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "Janus test failure vs a dependency version bump issue; not the same bug." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "Config save/load losing token IDs vs tokenizer AttributeError on tasks; different failure modes." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Hidden-states computation bug vs incorrect visual encoder key serialization; unrelated Qwen3.5 issues." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Missing Qwen2.5-VL export/import symbol vs Float8 storage lookup error; distinct API/runtime problems." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Tensor-parallel OOM vs unreleased minimum PEFT version bump; unrelated." + }, + { + "left": "issue:46821", + "right": "issue:46832", + "accept": false, + "reason": "TimesFM window_size AttributeError vs RT-DETR feature-level mismatch; different model families and bugs." + }, + { + "left": "issue:45245", + "right": "issue:46710", + "accept": false, + "reason": "Category-limit runtime error vs post-upgrade accuracy regression; no shared defect." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing config field after 5.0 update vs init_empty_weights constructor argument mismatch; separate upgrade regressions." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_id on Qwen3-VL text config vs text_config unwrapping for composite causal LM configs; related family, but not the same concrete bug." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy field loading issue in BEiT vs device mismatch with CPU offload in bark-small; unrelated." + }, + { + "left": "issue:44589", + "right": "issue:46285", + "accept": false, + "reason": "Float8 storage type lookup vs CLIP text-model structure inconsistency; different subsystems." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both mention missing config fields, but one is pad_token_id on multimodal text config and the other is pad_token_idx in StableLmConfig; not the same bug." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Different Qwen3 MOE loading regressions with different triggers and errors." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id key warnings across many models vs CLIP text-model structure change; not one underlying issue." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "BEiT legacy-field load failure vs BigBirdTokenizer special-token registration bug; unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 28, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:34567", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:42490", + "issue:43065", + "issue:43493", + "issue:43606", + "issue:43696", + "issue:43701", + "issue:43844", + "issue:43873", + "issue:43937", + "issue:43957", + "issue:43994", + "issue:44112", + "issue:44423", + "issue:44464", + "issue:44493", + "issue:44554", + "issue:44734", + "issue:44869", + "issue:44964", + "issue:44977", + "issue:45072", + "issue:45084", + "issue:45200", + "issue:45229", + "issue:45446", + "issue:45520", + "issue:45588", + "issue:45823", + "issue:46050", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46832", + "issue:46846" + ], + "soft_pairs": [ + "issue:46399|issue:46846", + "issue:43606|issue:44112", + "issue:45072|issue:46832", + "issue:44464|issue:46424", + "issue:43873|issue:46050", + "issue:43844|issue:44964", + "issue:43994|issue:44493", + "issue:43937|issue:46701", + "issue:44869|issue:46752", + "issue:45588|issue:46050", + "issue:45200|issue:45229", + "issue:45229|issue:45823", + "issue:29942|issue:30333", + "issue:29159|issue:32090", + "issue:42490|issue:43696", + "issue:45084|issue:45520", + "issue:28218|issue:34567", + "issue:43065|issue:46369", + "issue:15354|issue:38175", + "issue:44554|issue:45446", + "issue:44977|issue:46701", + "issue:41720|issue:46710", + "issue:43493|issue:43701", + "issue:44423|issue:44734", + "issue:44977|issue:46710", + "issue:37428|issue:43957" + ], + "prepared_review_unit_hash": "fb940c46b71563f544980be370db330ca8e2c1d3fdcfc25a49bf11e9e3a36f00", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34913, + "estimated_input_tokens": 8729, + "estimated_eval_tokens": 17714 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The soft pairs are only broad thematic similarities; none look like the same underlying bug or change, so all candidate merges are rejected. This cluster is heterogeneous and should not be deduplicated into a single issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43873 is the best available anchor only because it is open and describes a broad runtime/offloading regression, but the cluster is too heterogeneous for true duplicate canonicalization.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the most reasonable triage anchor for the cluster, but it should not be treated as a duplicate magnet because the other issues span unrelated models and failure modes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46399", + "right": "issue:46846", + "accept": false, + "reason": "Different bugs: one is a missing decorator/registration crash, the other is missing TP-plan entries causing OOM/shape errors." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Unrelated: CPU offload device mismatch for Bark vs a GraniteSpeech CI test failure." + }, + { + "left": "issue:45072", + "right": "issue:46832", + "accept": false, + "reason": "Different code paths and models; dtype mismatch in inference is not the same as RTDetr forward shape handling." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward is distinct from cache max-length handling for quantized cache." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "One is quantization/offloading behavior, the other is broken save/load for sinq-quantized models." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "ZeRO-3 gradient growth during training is unrelated to loading a specific multimodal model on latest transformers." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 output corruption is a different symptom than the position-id key warning affecting many models." + }, + { + "left": "issue:43937", + "right": "issue:46701", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 is unrelated to DiffusionGemma pipeline failure." + }, + { + "left": "issue:44869", + "right": "issue:46752", + "accept": false, + "reason": "Whisper timestamp decoding crash and empty chat-template input crash are different functions and failure modes." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention sinkless-model AttributeError is unrelated to quantized model save/load corruption." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Text-only mm_token_type_ids defaulting is a different issue from multi-GPU inference OOM." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Both involve Gemma4, but one is inference OOM and the other is a device mismatch in forward; not the same bug." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures do not match MLFlow reporting after training-job failure." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency is unrelated to Trainer NoneType broadcast errors." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Local-path import warning has nothing to do with GPT-oss CUDA OOM." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template-node compilation error is unrelated to the flash_attn import KeyError on Python 3.13." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer added-space behavior and TrainerState token counter updates are different subsystems and bugs." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as memory overflow in sam3-litetext." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure is unrelated to SigLIP2 zero-probability outputs." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness on mismatched head dims is distinct from an incorrect PyTorch version gate." + }, + { + "left": "issue:44977", + "right": "issue:46701", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug is unrelated to DiffusionGemma pipeline failure." + }, + { + "left": "issue:41720", + "right": "issue:46710", + "accept": false, + "reason": "Qwen3 auto device-map CUDA assert and DeepSeek accuracy regression are different problems and likely different code paths." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 JAX/HF discrepancy is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing crash are both serve bugs but not the same underlying defect." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash attention utility import error is unrelated to meta-device model loading breakage." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 29, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28218", + "issue:29870", + "issue:29942", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:43232", + "issue:43278", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43824", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43975", + "issue:44186", + "issue:44466", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44737", + "issue:44849", + "issue:44987", + "issue:45072", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45356", + "issue:45381", + "issue:45406", + "issue:45520", + "issue:45588", + "issue:45759", + "issue:45823", + "issue:46121", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46821", + "issue:46894" + ], + "soft_pairs": [ + "issue:45406|issue:46121", + "issue:43295|issue:43450", + "issue:45072|issue:45198", + "issue:45588|issue:45759", + "issue:43540|issue:45200", + "issue:43295|issue:43873", + "issue:44186|issue:46192", + "issue:43572|issue:46821", + "issue:45823|issue:46192", + "issue:43295|issue:44610", + "issue:43550|issue:43844", + "issue:45200|issue:45823", + "issue:45520|issue:46178", + "issue:43450|issue:43575", + "issue:29870|issue:29942", + "issue:43572|issue:44987", + "issue:37428|issue:43824", + "issue:43232|issue:44849", + "issue:39692|issue:46192", + "issue:42490|issue:43493", + "issue:44521|issue:45290", + "issue:44554|issue:45356", + "issue:43295|issue:43493", + "issue:45381|issue:46209", + "issue:42915|issue:43824", + "issue:44466|issue:45127", + "issue:44737|issue:46894", + "issue:28218|issue:35141", + "issue:43824|issue:43975", + "issue:43278|issue:43856" + ], + "prepared_review_unit_hash": "41286fed517910d66e98a92f69332d114a6cb8f56938a2c97df940e47f92f21c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36936, + "estimated_input_tokens": 9234, + "estimated_eval_tokens": 18724 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft-edge pair looks like the same underlying bug/change; most are only loosely related by broad subsystems like tokenization, multimodal processors, or generation.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43295 is the strongest representative of the set: a concrete, user-facing regression in core processor/tokenizer handling for multimodal inputs, with clearer reproduction and broader impact than the other issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43295 is the best global issue candidate because it is a real regression with explicit breakage in a central API path, rather than a narrow model-specific or doc-only problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and a rope-validation TypeError are unrelated bugs." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Both involve multimodal processor code, but one is a processor.tokenizer regression and the other is a batched video-shape bug." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "Different models and different failures: inference dtype mismatch vs save_pretrained/tokenization failure." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "Flash-attention sink-less-model crash and composite-config text_config unwrapping are different code paths." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Different models and different symptoms; no shared underlying bug." + }, + { + "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": "Both are tokenizer-adjacent, but one crashes on LayoutLMv2 NER/padding and the other ignores return_tensors with text_kwargs." + }, + { + "left": "issue:43572", + "right": "issue:46821", + "accept": false, + "reason": "StableLmConfig pad_token_idx and TimesFM window_size are unrelated configuration errors." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device-map mismatch and return_tensors being ignored are different issues." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Both are multimodal processor/model mismatches, but they affect different model pipelines and different failure modes." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "torch.compile+SDPA failure is unrelated to ZeRO-3 gradient inflation." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Same model family, but one is a missing default tensor field and the other is a device placement bug." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Different import/backend registry failures with no shared bug path." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Video batching shape bug and tensor-parallel OOM are unrelated." + }, + { + "left": "issue:29870", + "right": "issue:29942", + "accept": false, + "reason": "Bart generate UnboundLocalError and Flash Attention 2 test failures are not the same bug." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Config pad_token_idx regression and loading failure for a specific model are different problems." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Both are ImportErrors, but they miss different symbols from different modules." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs sync issue and output_hidden_states behavior on Qwen3.5 are distinct." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "Documentation example errors are not the same as tokenizer argument handling." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Local-path import warning and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both touch apply_chat_template, but one is assistant-mask generation and the other is a crash on tool-call messages." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness and Kimi tokenizer regression are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "Processor/tokenizer regression and SigLIP2 HF-vs-JAX discrepancy are different issues." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Qwen2.5-VL video position IDs and FP8 inference on B200 are different bugs." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3Moe FP8 failure and a Qwen2.5-VL import error are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Both involve tied weights, but serialization inconsistency and LoRA-merging collapse are not the same underlying bug." + }, + { + "left": "issue:44737", + "right": "issue:46894", + "accept": false, + "reason": "XLNet CPU tensor creation and CohereASR loss shifting are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer spacing after added tokens and post_init reinitializing output embeddings are different bugs." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "ImportError for a missing class and DeepSeek detokenization regression are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Training/eval dtype drift and Qwen3 MoE memory usage are different problems." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 30, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28282", + "issue:33357", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:43262", + "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:43827", + "issue:43901", + "issue:43950", + "issue:44112", + "issue:44206", + "issue:44442", + "issue:44561", + "issue:44610", + "issue:44792", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45198", + "issue:45200", + "issue:45325", + "issue:45412", + "issue:45563", + "issue:45584", + "issue:45593", + "issue:45759", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46424", + "issue:46512", + "issue:46821", + "issue:46858", + "issue:46899", + "issue:46932" + ], + "soft_pairs": [ + "issue:43452|issue:46512", + "issue:43653|issue:44112", + "issue:43262|issue:43450", + "issue:43950|issue:46285", + "issue:43334|issue:43525", + "issue:45200|issue:46899", + "issue:44442|issue:45198", + "issue:45200|issue:46932", + "issue:44898|issue:45412", + "issue:43526|issue:46192", + "issue:44792|issue:45563", + "issue:45200|issue:45325", + "issue:43819|issue:44112", + "issue:41720|issue:46209", + "issue:43335|issue:43550", + "issue:44561|issue:46234", + "issue:43827|issue:43901", + "issue:44610|issue:44857", + "issue:43819|issue:45083", + "issue:43388|issue:43526", + "issue:43606|issue:45083", + "issue:44857|issue:45593", + "issue:46512|issue:46821", + "issue:44206|issue:45584", + "issue:38175|issue:45412", + "issue:45759|issue:46285", + "issue:46424|issue:46858", + "issue:43335|issue:43653", + "issue:37428|issue:43425", + "issue:28282|issue:33357" + ], + "prepared_review_unit_hash": "534f0b749d3c0e7e0129bcc58ec4e4248d6501869d6d9af2456cdc46f9e5063b", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37093, + "estimated_input_tokens": 9274, + "estimated_eval_tokens": 18804 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set does not form a coherent duplicate cluster: the issues span unrelated bugs across loading, processors, generation, caching, and model-specific regressions. No soft edge looks like the same underlying defect, so the cluster should be split rather than merged.", + "confidence": 0.93, + "canonical_issue_reason": "Issue #43950 is the broadest core regression in the set: a from_pretrained loading corruption bug that can affect many models and code paths, making it the best representative if one issue must anchor the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "#43950 is the most central/high-impact issue among mostly model-specific reports, and its scope is wide enough to represent the cluster\u2019s loader/regression theme.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "Different failures: gguf loading path vs stale config dtype after from_config; no shared underlying bug." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Tokenizer special-token registration and a CI device-override failure are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate default vs batched video shape bug; different subsystems and causes." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Both touch loading/model structure, but one is non-persistent buffer corruption and the other is CLIP class API inconsistency." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Similar symptom name, but different model configs and likely separate fixes." + }, + { + "left": "issue:45200", + "right": "issue:46899", + "accept": false, + "reason": "Gemma4 text-only token-type-id defaulting and Gemma4 quantized embedder dtype casting are distinct bugs." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Two separate speech-tokenizer failures for different models and code paths." + }, + { + "left": "issue:45200", + "right": "issue:46932", + "accept": false, + "reason": "The later Gemma4 issue explicitly references another fix; it is not the same as the token_type_ids bug." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Image-size/interpolation failure and RT-DETR memory cleanup are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "Processor label reduction bug vs return_tensors being ignored; different API layers and symptoms." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "Janus image generation test failure is not the same as a stale warning in paged generate()." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal issues, but one is token-type IDs and the other is rope position scaling; different defects." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC latent reconstruction mismatch and a GraniteSpeech CI test issue are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Different failure modes: A800 cudaErrorAssert during auto device mapping vs FP8 inference failure on B200." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug and Bamba torch.compile+SDPA failure do not share a code-path." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Missing import_utils helper after v5 removal vs a different missing availability helper; not the same bug." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs-related, but one is pipeline removal references and the other is text-classification return_all_scores behavior." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "OmDet processor input-size mismatch and LwDetr AMP crash are separate model-specific bugs." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "DAC latents mismatch and qwen3_omni_moe length helper behavior are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both involve label handling, but one is gather_for_metrics truncation and the other is BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch and an audio length helper bug are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Both are vision-related, but one is AMP instability and the other is missing auxiliary losses when denoising is off." + }, + { + "left": "issue:46512", + "right": "issue:46821", + "accept": false, + "reason": "Stale config dtype after from_config is not the same as TimesFM window_size attribute errors." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Audio feature-extractor center-arg crash and Whisper empty-transcription failure after token alignment are different bugs." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and RT-DETR memory retention are unrelated." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Composite Qwen CausalLM config unwrapping vs CLIP text-model structure mismatch are different model-API issues." + }, + { + "left": "issue:46424", + "right": "issue:46858", + "accept": false, + "reason": "Cache length config limitation and static-cache slice arithmetic on non-CUDA backends are separate cache bugs." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Sparse-layer config creation and tokenizer special-token registration are unrelated." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Missing internal flash-attention symbol and broad Torch 2.10 incompatibility are not the same concrete defect." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "PyTorch-missing ImportError and a MacOS bus error with CLIP are completely 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": 26, + "review_unit_index": 31, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:26413", + "issue:28218", + "issue:29127", + "issue:29942", + "issue:31501", + "issue:32090", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43720", + "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:44779", + "issue:44855", + "issue:44857", + "issue:45083", + "issue:45290", + "issue:45310", + "issue:45478", + "issue:45520", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:46170", + "issue:46326", + "issue:46431", + "issue:46566" + ], + "soft_pairs": [ + "issue:29942|issue:32090", + "issue:43611|issue:44466", + "issue:43335|issue:43504", + "issue:44855|issue:45083", + "issue:43550|issue:44112", + "issue:45520|issue:46170", + "issue:44448|issue:44743", + "issue:45290|issue:46326", + "issue:37428|issue:44351", + "issue:44514|issue:44521", + "issue:43335|issue:45083", + "issue:43335|issue:45593", + "issue:43335|issue:44112", + "issue:43278|issue:43825", + "issue:29127|issue:32090", + "issue:45588|issue:45859", + "issue:42490|issue:43854", + "issue:34689|issue:35141", + "issue:43819|issue:44493", + "issue:46431|issue:46566", + "issue:39692|issue:43526", + "issue:43720|issue:44112", + "issue:43606|issue:44857", + "issue:45310|issue:45478", + "issue:44448|issue:44779", + "issue:39692|issue:44493", + "issue:26413|issue:31501", + "issue:28218|issue:32090", + "issue:44206|issue:45823" + ], + "prepared_review_unit_hash": "99053abad02393e4ab786e8cbe6d34ddf14d2f0ce2d7f2dd02395efa3e24dc59", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36073, + "estimated_input_tokens": 9019, + "estimated_eval_tokens": 18294 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: it contains many unrelated bug reports, with only one clear duplicate pair. The only strong merge candidate is the Qwen3.5 MoE from_pretrained regression (45310/45478).", + "confidence": 0.93, + "canonical_issue_reason": "issue:45310 is the best canonical anchor because issue:45478 is a near-exact duplicate of the same Qwen3.5 MoE from_pretrained failure, and 45310 is the earlier report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the strongest issue candidate in this set because it has an exact duplicate (45478) and a clear, concrete failure mode; most other items are unrelated to each other.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Different failures: Flash Attention 2 tests vs a Trainer _gpu_broadcast_one NoneType error." + }, + { + "left": "issue:43611", + "right": "issue:44466", + "accept": false, + "reason": "Both involve loading/serialization regressions, but they affect different code paths and symptoms." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "Unrelated models and bugs: SwitchTransformers sparse-layer config vs BEiT legacy-field loading." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Different modules and problems: DebertaV2 import parsing vs qwen3_omni_moe helper behavior." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Different subsystems: torch.compile/SDPA model failure vs a CI test stale device override." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Different import/dependency failures on different platforms and libraries." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Different model behaviors: Pegasus v4/v5 output drift vs recurrent state reset in modular_qwen3_5." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both touch chat templating, but one is an assistant-message/tool-call crash and the other is a mistral_common continue_final_message bug." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and unrelated fixes." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Same broad API area, but one is a batching/padding crash and the other is incorrect assistant masks." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "Different model/config logic; no shared underlying defect." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model bugs: SwitchTransformers config vs D-FINE auxiliary loss handling." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "No shared code-path problem; one is model config creation, the other is a CI device-override test." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Different behaviors: dtype mismatch during evaluate vs a pipeline error message regression." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "Different areas and symptoms; error-message clarity vs distributed broadcast type error." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Both involve offload/attention-related code, but the concrete failures are different and not the same bug." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "Different issues: trailing-slash import warning vs unit-test model loading failure." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Different loading/weight-tie bugs affecting different models and mechanisms." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "Different symptoms and code paths: DAC latent consistency vs unexpected position-id keys." + }, + { + "left": "issue:46431", + "right": "issue:46566", + "accept": false, + "reason": "A missing chat template error is unrelated to DiffusionGemma meta-device loading." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "Different model families and bugs: SigLIP2 docs example vs BEiT fast image processor label reduction." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading bug is unrelated to a GraniteSpeech CI test failure." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "Both are runtime crashes, but one is CPU-offload device mismatch and the other is AMP/CUDA loss failure." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same title, same Qwen3.5 MoE from_pretrained regression, with 45478 clearly duplicating 45310." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both are regression reports, but they affect different models/tokenizers and different failure modes." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "No shared underlying bug; documentation/model mismatch is unrelated to position-id key warnings." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Training startup/resume issues are too generic here; the concrete failure reports do not match." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer spacing bug is unrelated to a Trainer distributed broadcast type error." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Different model/runtime failures with no shared concrete code path." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded overall. The summary stays conservative, and the one accepted soft-edge pair (45310/45478) is supported by identical titles and chronology. The other verdicts are mostly cautious false negatives given the limited evidence in titles alone. No obvious overstatement or mergeability issue stands out." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 32, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:27957", + "issue:29870", + "issue:29942", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:40444", + "issue:42175", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43381", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43701", + "issue:43927", + "issue:44360", + "issue:44448", + "issue:44493", + "issue:44561", + "issue:44617", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45127", + "issue:45229", + "issue:45245", + "issue:45290", + "issue:45381", + "issue:45459", + "issue:45593", + "issue:45698", + "issue:45920", + "issue:46192", + "issue:46234", + "issue:46399", + "issue:46655", + "issue:46697", + "issue:47000" + ], + "soft_pairs": [ + "issue:45127|issue:45245", + "issue:44617|issue:46697", + "issue:43701|issue:45083", + "issue:45459|issue:46234", + "issue:32090|issue:33357", + "issue:42490|issue:43575", + "issue:44857|issue:45072", + "issue:45593|issue:47000", + "issue:43550|issue:45083", + "issue:38175|issue:44493", + "issue:27957|issue:34689", + "issue:29942|issue:35141", + "issue:45290|issue:45381", + "issue:29870|issue:32090", + "issue:10000|issue:15354", + "issue:43278|issue:43329", + "issue:44448|issue:44977", + "issue:45290|issue:46655", + "issue:44360|issue:44561", + "issue:44855|issue:46399", + "issue:45920|issue:46192", + "issue:29870|issue:33290", + "issue:37428|issue:42175", + "issue:40444|issue:43526", + "issue:44977|issue:45229", + "issue:43278|issue:43381", + "issue:43927|issue:45698", + "issue:32090|issue:33290" + ], + "prepared_review_unit_hash": "ff23fc159436158cf0ee3c3d9a76a655538301c7c497469c33903b931c3c40c8", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36206, + "estimated_input_tokens": 9052, + "estimated_eval_tokens": 18360 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are broadly heterogeneous across models, processors, generation, loading, and training paths. I do not see any soft edge pair that looks like the same underlying bug/change, so all soft links should be rejected.", + "confidence": 0.96, + "canonical_issue_reason": "Issue:45920 is the strongest anchor: it reports a broad, concrete tokenizer regression affecting multiple models, with clear user impact and higher duplicate-likelihood than the more model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:45920 is the best overall representative because it is high-signal, broadly scoped, and plausibly central to nearby tokenizer-related reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "Different bug classes: LoRA/vocab merge collapse vs category-cardinality runtime error." + }, + { + "left": "issue:44617", + "right": "issue:46697", + "accept": false, + "reason": "Unrelated failures: video-model OOM vs image-processor over-padding." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch is unrelated to a multimodal feature-length helper bug." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Both touch import/utilities, but one is tokenizer error masking and the other is a missing import_utils symbol." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Distributed broadcast NoneType error is not the same as a MacOS bus error on CLIP loading." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Local-path slash warning and tensor-parallel OOM are different load-path failures." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype/mixed precision, but they affect different code paths and symptoms." + }, + { + "left": "issue:45593", + "right": "issue:47000", + "accept": false, + "reason": "Different detection-loss bugs: missing auxiliary losses vs infeasible cost-matrix crash." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure is unrelated to a Qwen3 omni feature-length helper bug." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Zero probabilities in one model do not match a broad position-id key mismatch issue." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Both are model-specific loading issues, but they involve different models and failures." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to embedding reinitialization after resize." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template crashes with tool calls are unrelated to Qwen2.5-VL video position-id errors." + }, + { + "left": "issue:29870", + "right": "issue:32090", + "accept": false, + "reason": "Both are runtime errors, but one is a BART generate-path bug and the other is a trainer broadcast type error." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "Different frameworks and failures: DistilBert issue vs torch.jit.script GeneratorExp export error." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift is not the same as undefined multimodal-token helper names in the video branch." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Sequence-to-sequence output regression vs flash-attention generation failure are distinct problems." + }, + { + "left": "issue:45290", + "right": "issue:46655", + "accept": false, + "reason": "Chat-template tool-call handling is unrelated to video-processor padding calculation." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "A missing ReLU in a discussion/implementation issue is unrelated to removal of a utility function breaking trust_remote_code models." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Both mention decorators, but one is a Python 3.13 import/parser issue and the other is a registration/value-error in Qwen3.5 kernels." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Tokenizer ID regression is different from return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:29870", + "right": "issue:33290", + "accept": false, + "reason": "Generate-path decoder_attention_mask crash is unrelated to deepspeed Adafactor OOM." + }, + { + "left": "issue:37428", + "right": "issue:42175", + "accept": false, + "reason": "Flash-attention import failure and missing TensorFlow backend dependency are unrelated." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Multi-image VL fine-tuning failure is not the same as BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Flash-attention generation bug and multi-GPU OOM are different failure modes." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Eval dtype mismatch is unrelated to gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load behavior, but one is custom token IDs in config and the other is wrong custom-module resolution." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast NoneType error is not the same as optimizer memory exhaustion." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 33, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:25251", + "issue:29127", + "issue:29870", + "issue:29942", + "issue:30064", + "issue:32090", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43278", + "issue:43381", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43873", + "issue:43976", + "issue:44186", + "issue:44206", + "issue:44361", + "issue:44442", + "issue:44448", + "issue:44466", + "issue:44479", + "issue:44561", + "issue:44568", + "issue:44743", + "issue:44779", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45229", + "issue:45372", + "issue:45440", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45803", + "issue:45823", + "issue:46121", + "issue:46421", + "issue:46858" + ], + "soft_pairs": [ + "issue:46421|issue:46858", + "issue:43873|issue:45229", + "issue:42490|issue:43873", + "issue:43976|issue:45440", + "issue:44448|issue:44466", + "issue:25251|issue:32090", + "issue:44568|issue:45491", + "issue:44871|issue:44977", + "issue:29127|issue:34567", + "issue:44561|issue:44779", + "issue:43381|issue:43701", + "issue:33357|issue:39692", + "issue:45588|issue:45823", + "issue:45198|issue:45803", + "issue:43278|issue:43575", + "issue:41720|issue:45229", + "issue:44977|issue:45563", + "issue:44186|issue:45198", + "issue:43611|issue:46121", + "issue:44361|issue:45198", + "issue:29870|issue:34689", + "issue:43653|issue:44743", + "issue:44206|issue:44479", + "issue:29942|issue:30064", + "issue:42915|issue:42994", + "issue:43756|issue:45440", + "issue:43653|issue:44442", + "issue:45372|issue:45588" + ], + "prepared_review_unit_hash": "00df70f7f1931bce5d3f779f609f340618c4ee9ec0c7d70c5184cdfa0d4b226a", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36287, + "estimated_input_tokens": 9072, + "estimated_eval_tokens": 18400 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a noisy mix of unrelated Transformers issues; the soft links are theme-level similarities, not true duplicates. No soft edge looks like the same underlying bug/change.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43873 is the closest thing to a hub because it has one of the few multiple soft links and sits in the broader quantization/offloading area, but even that match is weak and the cluster is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best representative only in a loose thematic sense; it is not a strong duplicate canonical because the linked issues are mostly about different failures.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache/prefill/compile behavior, but one is about lazy cache init causing recompilation while the other is a device-tensor seq-length bug in static cache arithmetic." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Quantization/offloading vs multi-GPU CUDA OOM are different failure modes and likely different root causes." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Trailing-slash local model import warning is unrelated to quantization offloading behavior." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python version compatibility issue and DeepSeekV3MoE divergence are unrelated." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Both are v5 regressions, but one is output drift for Pegasus and the other is serialization of tied weights; different code paths." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Top-k pipeline shape change and Trainer broadcast NoneType are unrelated bugs." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer special-token handling and Gemma3 mixed-length NaN embeddings are different subsystems and symptoms." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id config mismatch and Qwen3.5 flash-attention generation failure are not the same bug." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and TrainerState token-count update are unrelated." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Removal of is_torch_fx_available breaking remote-code models is distinct from Deepseek tokenizer correctness." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Checkpointing in eval mode and resume_from_checkpoint key mismatch are different training-state issues." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error on CLIP loading and SigLIP2 doc example errors are separate problems." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Flash-attention s_aux=None crash and Gemma4 PLE device mismatch are different forward-path bugs." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure and Fuyu bare-except swallowing errors are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Embedding dtype drift in eval and tensor-parallel OOM on Qwen2-57B are not the same underlying issue." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Both involve large-model multi-GPU/device mapping, but one is a cudaErrorAssert on Qwen3 and the other is OOM on Gemma4." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation failure and stale num_return_sequences warning are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes and Wav2Vec2 save/tokenization failure are different tokenizer/data-path bugs." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "base_model_prefix loading regression and rope-parameter dict conversion TypeError are unrelated." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLuke tokenizer AttributeError and Wav2Vec2 save/tokenization failure do not share a concrete root cause." + }, + { + "left": "issue:29870", + "right": "issue:34689", + "accept": false, + "reason": "BART generate decoder_attention_mask bug and Llama 3.2 vision loading breakage are different model-specific regressions." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and Qwen3_5 recurrent-state reset are unrelated." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "LasrFeatureExtractor center-arg crash and Qwen video-input regression are different code paths." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures and void segmentation-map processing are only both test/vision-related at a high level, not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both involve quantization, but one is FineGrainedFP8Config load failure and the other is quantized model saving; different operations." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Smollm3 RoPE-layer mismatch and DeepSeekV3MoE divergence are unrelated architecture issues." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer mask-token bug and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer failures." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 processor import failure and flash-attention s_aux=None crash are unrelated." + }, + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache behavior, but the first is about recompilation from lazy init while the second is about device-tensor seq-length breaking slice arithmetic 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": 26, + "review_unit_index": 34, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:42490", + "issue:42915", + "issue:43065", + "issue:43066", + "issue:43381", + "issue:43412", + "issue:43421", + "issue:43526", + "issue:43575", + "issue:43701", + "issue:43824", + "issue:43931", + "issue:43994", + "issue:44186", + "issue:44265", + "issue:44279", + "issue:44387", + "issue:44442", + "issue:44462", + "issue:44464", + "issue:44857", + "issue:45072", + "issue:45198", + "issue:45245", + "issue:45250", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45712", + "issue:45834", + "issue:46369", + "issue:46640", + "issue:46693", + "issue:46701" + ], + "soft_pairs": [ + "issue:43065|issue:43994", + "issue:43065|issue:45412", + "issue:43412|issue:45588", + "issue:44265|issue:45198", + "issue:45491|issue:46693", + "issue:29127|issue:33357", + "issue:43575|issue:43701", + "issue:45245|issue:45563", + "issue:45459|issue:45712", + "issue:42490|issue:43381", + "issue:29127|issue:35141", + "issue:45440|issue:45593", + "issue:45372|issue:45406", + "issue:43994|issue:45412", + "issue:43421|issue:45491", + "issue:45406|issue:46701", + "issue:45412|issue:46369", + "issue:43066|issue:44462", + "issue:45372|issue:45459", + "issue:43575|issue:44387", + "issue:44442|issue:45072", + "issue:44464|issue:45584", + "issue:43526|issue:45834", + "issue:43526|issue:46693", + "issue:27957|issue:33290", + "issue:43824|issue:44279", + "issue:27957|issue:29127", + "issue:43994|issue:46369", + "issue:42915|issue:43931", + "issue:33290|issue:35141", + "issue:33290|issue:34689", + "issue:44857|issue:45588", + "issue:43526|issue:44186", + "issue:45250|issue:46640" + ], + "prepared_review_unit_hash": "983fe489bfd50d880e21e3148c77cc512524af79559e2fba2faf4bb9eddb8e91", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37089, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a broad mix of unrelated Transformers issues across model loading, tokenizers, quantization, export, memory, and attention. None of the soft pairs look like the same underlying bug with enough specificity to merge.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Different models and symptoms: Sam3PixelDecoder dummy Conv2d vs Siglip2 nonsensical outputs. No shared code-path evidence." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Sam3 pixel-decoder placeholder layer and RT-DETR GC memory leak are unrelated bugs." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Gemma3n export failure and flash_attention.py s_aux=None crash are different failure modes in different paths." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export with torch_compilable_check is unrelated to Wav2Vec2 save_pretrained/tokenization failure." + }, + { + "left": "issue:45491", + "right": "issue:46693", + "accept": false, + "reason": "Both mention FA2, but one is NaN embeddings from sliding-window padding and the other is a performance regression in long-context generation." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 box validation messaging and CLIP bus error on macOS have no common underlying bug." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "Tensor-parallel OOM on Qwen2-57B is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-count limit error and stale num_return_sequences warning are different issues in different code paths." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Tokenizer error masking when protobuf is missing is unrelated to leftover dummy classes leaking without torch." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-path import warning and gradient-checkpointing-in-eval behavior are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Error-message clarity for LayoutLMv3 boxes is unrelated to embedding reinitialization after resize." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3MoE parity divergence and D-FINE auxiliary-loss omission are different model-specific bugs." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Importing ReasoningEffort from mistral_common and Gemma4Processor missing _tokenizer in serve are distinct problems." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "Siglip2 inference quality issue and RT-DETR memory leak are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime special-token post-processor update bug is unrelated to Gemma3 sliding-window NaNs." + }, + { + "left": "issue:45406", + "right": "issue:46701", + "accept": false, + "reason": "Gemma4 serve attribute error and DiffusionGemma image-text-to-text pipeline failure are different processors/code paths." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "RT-DETR GC memory leak and sam3-litetext memory overflow are not the same issue." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong decoder type in Transformers v5 and AutoTokenizer ignoring tokenizer.json are separate tokenizer-loading bugs." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Dependency import failure for Gemma4 and protobuf-masked tokenizer errors are unrelated." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both are OOM-related, but one is TP load OOM for Qwen2-57B and the other is increased CUDA reserved memory under int4 quantization." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer loading and bfloat16 dtype mismatches are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency and Whisper empty-transcription failure after align_special_tokens do not share a code-path." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and Kosmos2.5 long-OCR index error are different preprocessing issues." + }, + { + "left": "issue:43526", + "right": "issue:46693", + "accept": false, + "reason": "No meaningful overlap: image-label reduction bug vs FA2 long-context generation regression." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and adafactor/deepspeed OOM are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Specific Qwen2_5_VL import error is not the same as a generic dependency issue report." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "Flash Attention support for XLMRoberta and LayoutLMv3 box error messaging are unrelated." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Optimizer/deepspeed OOM and embedding reinitialization after resize are different bugs." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Adafactor/deepspeed OOM and Llama 3.2 90B Vision loading breakage are not the same issue." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetr float16 AMP crash and flash_attention.py s_aux=None crash are distinct failures." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug and LayoutLMv2Tokenizer NER/padding crash are unrelated." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Both are generic FA2 titles, but there is no evidence they describe the same concrete 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": 1, + "component_count": 26, + "review_unit_index": 35, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:27957", + "issue:29870", + "issue:42175", + "issue:42490", + "issue:43262", + "issue:43316", + "issue:43334", + "issue:43381", + "issue:43493", + "issue:43525", + "issue:43531", + "issue:43643", + "issue:43646", + "issue:43720", + "issue:43749", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44206", + "issue:44442", + "issue:44479", + "issue:44514", + "issue:44743", + "issue:44811", + "issue:44855", + "issue:44871", + "issue:44898", + "issue:44913", + "issue:44977", + "issue:45003", + "issue:45137", + "issue:45245", + "issue:45325", + "issue:45381", + "issue:45440", + "issue:45588", + "issue:45759", + "issue:45859", + "issue:45907", + "issue:46121", + "issue:46234", + "issue:46399", + "issue:46697", + "issue:46821" + ], + "soft_pairs": [ + "issue:43334|issue:43531", + "issue:46121|issue:46399", + "issue:43381|issue:43873", + "issue:45588|issue:46821", + "issue:44514|issue:45381", + "issue:43646|issue:44077", + "issue:44977|issue:45245", + "issue:43761|issue:45859", + "issue:44743|issue:44811", + "issue:43262|issue:44206", + "issue:44442|issue:44898", + "issue:42490|issue:44743", + "issue:42175|issue:46234", + "issue:45003|issue:45440", + "issue:43720|issue:44442", + "issue:27957|issue:29870", + "issue:43316|issue:43525", + "issue:44077|issue:44479", + "issue:43493|issue:43827", + "issue:44062|issue:45907", + "issue:45325|issue:46697", + "issue:43873|issue:44871", + "issue:44855|issue:45137", + "issue:43643|issue:44913", + "issue:43316|issue:45759", + "issue:43749|issue:43931" + ], + "prepared_review_unit_hash": "9bec96e22763cdbea3c2219448300edd6e94a1d967bad0d130f7fba4b5a2286e", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35122, + "estimated_input_tokens": 8781, + "estimated_eval_tokens": 17818 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is a heterogeneous mix of unrelated Transformer bugs. None of the soft pairs look like the same underlying defect; most only share a broad subsystem or model family name, which is not enough to merge them as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43873 is the strongest standalone anchor in the set because it has the most discussion and inbound references, but the overall collection is too diverse to form a true duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the most substantively triaged and highest-activity issue here, so it is the best representative if a single issue must be chosen.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Both are Qwen-related, but one is a missing pad_token_id config error and the other is a sliding_window behavior issue; different code paths and symptoms." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "Rope validation TypeError and Qwen3.5 kernel registration failure are unrelated bugs despite both touching rotary/rope-adjacent code." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and quantization offloading are separate behaviors with no concrete shared failure mode." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "Flash-attention sink handling and TimesFM window_size AttributeError are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both involve Qwen2.5-VL, but one is batched chat-template padding and the other is video vision_position_ids; not the same bug." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Custom model init regression and patchtsmixer post_init validation are different initialization issues, not one mergeable fix." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3.5 flash-attention generation and category cardinality limit errors are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and Qwen3.5Moe CPU offload incompatibility affect different models and mechanisms." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Recurrent cache reset in Qwen3.5 modular code and Whisper batch_decode skipping special tokens are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and LASR feature extractor center-arg crash are different audio-processing bugs." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer load failure for FastSpeech2Conformer and Perceiver image classification padding/resizing are unrelated." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local path import warnings and Qwen3.5 recurrent-state reset are not the same underlying issue." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Torch install backend packaging and missing import_utils availability helper are unrelated packaging/API issues." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "Unsafe sys.modules access and DeepseekV3MoE divergence are distinct implementation problems." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight loading and tokenizer loading for FastSpeech2Conformer are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:29870", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and BART generate() decoder_attention_mask error are unrelated model-specific bugs." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Both are config/API shape complaints, but Gemma3TextConfig inconsistency and Llama4Config missing pad_token_id are different concrete failures." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "patchtsmixer post_init deprecation and Qwen video-input regression are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation drift and docs still referencing pipeline() are completely separate concerns." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "AddedToken keyword duplication and list_repo_files revision forwarding are unrelated API bugs." + }, + { + "left": "issue:45325", + "right": "issue:46697", + "accept": false, + "reason": "Qwen2.5-VL rope position scaling and Swin2SR image over-padding are unrelated image/video preprocessing issues." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading failure and Gemma-3 eos_token_id inconsistency do not share a concrete code-path bug." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 torch.jit parsing error and DeepSpeed ZeRO3 deque pop error are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "trust_remote_code missing fields and GPTNeoXConfig rotary_pct reset are both config-related, but they are different persistence/loading bugs." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Composite-config API discrepancy and AutoModelForCausalLM.from_config text_config unwrapping are related only at a high level, not the same bug." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU-efficient loading breakage and Qwen3-VL weight-shape mismatch are separate loading failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 36, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43452", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43761", + "issue:43825", + "issue:43931", + "issue:44077", + "issue:44079", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44464", + "issue:44496", + "issue:44534", + "issue:44871", + "issue:44913", + "issue:44987", + "issue:45216", + "issue:45584", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46178", + "issue:46234", + "issue:46291", + "issue:46365", + "issue:46421", + "issue:46506", + "issue:46821", + "issue:46829", + "issue:46932" + ], + "soft_pairs": [ + "issue:44987|issue:46178", + "issue:44190|issue:44496", + "issue:43232|issue:43388", + "issue:44442|issue:45803", + "issue:43931|issue:45216", + "issue:43761|issue:44079", + "issue:43701|issue:46506", + "issue:43761|issue:44188", + "issue:46829|issue:46932", + "issue:42617|issue:43575", + "issue:43550|issue:43701", + "issue:43575|issue:43742", + "issue:27957|issue:32090", + "issue:43606|issue:43701", + "issue:25251|issue:28282", + "issue:42907|issue:45698", + "issue:43295|issue:43825", + "issue:16998|issue:22355", + "issue:43295|issue:44871", + "issue:44871|issue:45584", + "issue:46234|issue:46291", + "issue:44464|issue:46421", + "issue:43316|issue:43716", + "issue:44077|issue:44534", + "issue:43316|issue:46821", + "issue:43646|issue:44913", + "issue:44871|issue:46050", + "issue:43452|issue:44496", + "issue:45803|issue:46365" + ], + "prepared_review_unit_hash": "0d368db69fa49ab5fb8fbbfbd3d4780d0f40bf1d254ce9bd02aac909a2fd9914", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36303, + "estimated_input_tokens": 9076, + "estimated_eval_tokens": 18408 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate edges are mostly false positives: they share broad Transformers themes, but each pair points to a different concrete failure mode, model, or code path. No soft edge is safe to merge as a duplicate.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45698 is the broadest and most discussed save/load regression in the set, so it is the least-bad anchor issue; however, the overall cluster is heterogeneous and not a true duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45698 is the best representative issue by scope and discussion activity, but only as a loose anchor\u2014not as a duplicate canonical for the rest of the items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Different bugs: model-loading failure on a specific repo/version vs a missing backend registration for tensorflow_text." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Different failure classes: local dataset loading in a training script vs unrecognized model/config missing model_type." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs sync bug after sync_gpus vs gather_for_metrics dropping tuple labels in the last batch." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer import/load failure vs a bare except swallowing interrupts and masking errors; unrelated code paths." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "One is a weight-shape mismatch when loading, the other is a save_pretrained regression for Qwen3.5 checkpoints." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "Hidden-states output regression in CLIPVisionModel vs ModelOutput key assignment logic; different layers of the stack." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "Checkpoint resume key mismatch vs CPU/CUDA numerical divergence; not the same underlying bug." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIP hidden_states output bug vs attention-kernel divergence under torch.compile; different symptoms and fixes." + }, + { + "left": "issue:46829", + "right": "issue:46932", + "accept": false, + "reason": "VideoPrism output-field bug vs Gemma4 quantized embedder dtype casting; different models and code paths." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "Generic 3D parallel script failure vs OOM when loading a specific Qwen2 model with tensor parallelism." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba vs resume_from_checkpoint key mismatch; unrelated runtime behaviors." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "OOM during tp load of a large model vs key error loading MobileLLM-125M; different causes." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 support issue for XLMRoberta vs Trainer GPU broadcast NoneType error; no shared bug." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch vs checkpoint resume key mismatch; different subsystems entirely." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline top_k output nesting bug vs missing PyTorch import error; unrelated." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load behavior, but one is quantized-model saving after dequantization and the other is custom-module restoration; not the same bug." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "Processor/tokenizer regression with images vs an error-message wording bug for unsupported translation tasks." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "Question about model_max_length vs missing transformers.onnx module import; no duplicate relationship." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Different issues: custom model code breaking around processor.tokenizer/images vs eos_token_id config inconsistency." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id mismatch vs Whisper empty-transcription generation failure after align_special_tokens; different models and failure points." + }, + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Missing import_utils helper vs pipelines import requiring kernels version/revision; unrelated dependency issues." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Chunked generation output inconsistency under compiled forward vs spurious recompilation from lazy cache init; same area but not the same bug." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API mismatch vs Mistral-3 image-preprocessor dtype mismatch; different model families and fixes." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init policy for patchtsmixer vs non-persistent buffer junk in v5; unrelated issues." + }, + { + "left": "issue:43316", + "right": "issue:46821", + "accept": false, + "reason": "Gemma3TextConfig discrepancy vs TimesFM window_size AttributeError; different APIs and models." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Custom model initialization breakage in Transformers 5 vs GPTNeoX rotary_pct not persisting on reload; different bugs." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "eos_token_id configuration mismatch vs quantized sinq save_pretrained/from_pretrained breakage; both serialization-adjacent but not duplicates." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "gguf_file loading bug for tokenizer/model vs missing model_type in config.json; separate loading failures." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "Bare except in Fuyu tensor conversion vs Trainer import failing until datasets is uninstalled; unrelated import/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": 26, + "review_unit_index": 37, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:35141", + "issue:41950", + "issue:42490", + "issue:43065", + "issue:43352", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43475", + "issue:43504", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43572", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:44162", + "issue:44190", + "issue:44242", + "issue:44514", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45200", + "issue:45406", + "issue:45542", + "issue:45588", + "issue:45600", + "issue:45632", + "issue:45698", + "issue:45803", + "issue:46170", + "issue:46178", + "issue:46202", + "issue:46234", + "issue:46369", + "issue:46424", + "issue:46640", + "issue:46894" + ], + "soft_pairs": [ + "issue:44162|issue:44792", + "issue:43404|issue:43441", + "issue:44242|issue:46894", + "issue:45632|issue:46424", + "issue:44610|issue:46369", + "issue:43540|issue:44514", + "issue:43065|issue:44617", + "issue:45542|issue:46234", + "issue:45803|issue:46170", + "issue:43650|issue:46202", + "issue:43572|issue:46178", + "issue:43504|issue:45698", + "issue:44792|issue:45600", + "issue:44792|issue:44977", + "issue:45127|issue:45200", + "issue:43352|issue:46640", + "issue:43673|issue:44743", + "issue:41950|issue:43475", + "issue:43388|issue:43688", + "issue:43525|issue:43531", + "issue:44534|issue:44913", + "issue:35141|issue:42490", + "issue:44190|issue:44964", + "issue:44734|issue:45406", + "issue:44855|issue:45588" + ], + "prepared_review_unit_hash": "258729f45e4d05d1b91b42e942e994b72f1d846fa82d8b9110febf26b0d0a190", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 25, + "serialized_chars": 35154, + "estimated_input_tokens": 8789, + "estimated_eval_tokens": 17834 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous: most items are distinct model-, processor-, or config-specific bugs with only superficial wording overlap. I found no soft-edge pair that clearly describes the same underlying defect or change, so all candidate merges are rejected.", + "confidence": 0.91, + "canonical_issue_reason": "issue:44534 is the broadest cross-cutting regression in the set and most suitable as a representative issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44534 has the widest impact and a general Transformers v5 failure mode, making it the best single issue representative here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "Unrelated domains: ESM2 model breakage vs Janus image-generation test failure." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Both concern Mistral/Ministral models, but one is a tied-weight bug and the other is a Flash Attention support failure." + }, + { + "left": "issue:44242", + "right": "issue:46894", + "accept": false, + "reason": "Different loss bugs in different training setups; only a generic 'training loss' theme is shared." + }, + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "One is a trust_remote_code cache-path collision, the other is cache-length handling for quantized cache." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "Different SAM3 problems: processor/model image-size mismatch vs memory overflow." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Qwen3OmniMoe video-input processing vs Qwen2.5-VL batched chat templating; not the same bug." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Both mention SAM3 video/pixel decoder, but one is a dummy layer issue and the other is OOM behavior." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "Completely different backend/import issues with no shared code-path defect." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Bare except in tensor conversion is unrelated to Windows ROCm import failures." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "One is a generic placeholder issue, the other is a Gemma token roundtrip bug." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "StableLm config field regression vs backend mapping error; unrelated." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy preset-field loading bug vs wrong custom-module resolution after save/load." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "Janus test failure vs Sam3LiteText auto-mapping CI breakage; different root causes." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus image-generation test failure is unrelated to Qwen3.5 flash-attention generation issues." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA merged-vocab collapse with tied embeddings vs Gemma 4 token-type-id handling." + }, + { + "left": "issue:43352", + "right": "issue:46640", + "accept": false, + "reason": "Both mention Flash Attention, but one is a specific model support limitation and the other is a generic open issue." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Cache-related symptoms, but different mechanisms: GenerationMixin cache missing vs recurrent state reset in modular_qwen3_5." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Image-processor lookup in video classification is unrelated to SAM3 VisionEncoderOutput missing an attribute." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Different bugs in loss/metrics handling; no shared underlying defect." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Llama4 missing pad_token_id and Qwen3-MoE sliding_window behavior are separate config/model issues." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "General non-persistent buffer initialization bug vs GPTNeoX rotary_pct reload regression." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Token-embedding resize/post_init weight reinit vs local-model path trailing-slash import warning." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading in a training script vs loading a specific multimodal model." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation indexing crash vs Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 IndentationError from a scripted function vs flash_attention s_aux=None crash." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 38, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29870", + "issue:35141", + "issue:39692", + "issue:41720", + "issue:43065", + "issue:43335", + "issue:43550", + "issue:43644", + "issue:43673", + "issue:43827", + "issue:43854", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44479", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44792", + "issue:44855", + "issue:44869", + "issue:44877", + "issue:44913", + "issue:44933", + "issue:45072", + "issue:45245", + "issue:45362", + "issue:45381", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45632", + "issue:46082", + "issue:46369", + "issue:46421", + "issue:46617", + "issue:46655", + "issue:46697", + "issue:46832", + "issue:46858", + "issue:46897", + "issue:46932", + "issue:47013" + ], + "soft_pairs": [ + "issue:45072|issue:46932", + "issue:44869|issue:46832", + "issue:45245|issue:46858", + "issue:45632|issue:46617", + "issue:43673|issue:46421", + "issue:46082|issue:46697", + "issue:43827|issue:43994", + "issue:29870|issue:35141", + "issue:44617|issue:45584", + "issue:43931|issue:44060", + "issue:43644|issue:44913", + "issue:25251|issue:27957", + "issue:43335|issue:44855", + "issue:44242|issue:46897", + "issue:44933|issue:45362", + "issue:44077|issue:44877", + "issue:45381|issue:46369", + "issue:45563|issue:45584", + "issue:43550|issue:44190", + "issue:41720|issue:45245", + "issue:43065|issue:44610", + "issue:44479|issue:44734", + "issue:43335|issue:44190", + "issue:45588|issue:47013", + "issue:44617|issue:46369", + "issue:45381|issue:46655", + "issue:43854|issue:44792", + "issue:39692|issue:44190" + ], + "prepared_review_unit_hash": "b9c0ebeceb65d0c1a1a8b27aa79134c9748aa1c0f2095d4915a3ac2f8b08a40d", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36231, + "estimated_input_tokens": 9058, + "estimated_eval_tokens": 18372 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like a true duplicate; the overlaps are broad subsystem similarities rather than the same concrete bug. If one issue must serve as the anchor, issue:46858 is the clearest and most actionable report in the set.", + "confidence": 0.93, + "canonical_issue_reason": "issue:46858 is the most canonical-looking issue here: it has a precise failure mode, a specific trigger (static cache + torch.compile on non-CUDA backends), and an explicit root-cause description.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46858 is the best standalone issue because it is narrowly scoped and technically specific, making it easiest to triage and fix.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45072", + "right": "issue:46932", + "accept": false, + "reason": "Both mention dtype-related inference issues, but they affect different models and different code paths (bfloat16 inference vs quantized vision/audio embedders)." + }, + { + "left": "issue:44869", + "right": "issue:46832", + "accept": false, + "reason": "Whisper token-stream decoding crash and RT-DETR feature-level forward TypeError are unrelated failures in different components." + }, + { + "left": "issue:45245", + "right": "issue:46858", + "accept": false, + "reason": "One is a category-cardinality runtime limit, the other is a cache-length/device-tensor arithmetic bug; no shared underlying defect." + }, + { + "left": "issue:45632", + "right": "issue:46617", + "accept": false, + "reason": "Both involve trust_remote_code/cache behavior, but one is a leaf-directory path collision and the other is a FileNotFoundError while loading from cache." + }, + { + "left": "issue:43673", + "right": "issue:46421", + "accept": false, + "reason": "Both mention chunked prefill/cache, but one is missing cache state in v5.0.0 while the other is spurious recompilation from lazy initialization." + }, + { + "left": "issue:46082", + "right": "issue:46697", + "accept": false, + "reason": "These are unrelated config vs image-processor bugs; they do not share the same failing code path." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs referring to removed pipeline APIs is not the same as SigLIP2 producing bad outputs through AutoModel/pipeline." + }, + { + "left": "issue:29870", + "right": "issue:35141", + "accept": false, + "reason": "Bart generate() decoder mask initialization and output-embedding reinitialization after resize are distinct model bugs." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "A CUDA OOM in Sam3Video and Whisper empty-transcription generation failure are clearly different problems." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Model weight-shape mismatch on load and an incorrect tied-weights warning are different issues in Qwen3-family models." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk filling and GPTNeoX config reload losing rotary_pct are separate state/serialization bugs." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "Pipeline top_k nesting behavior and XLMRoberta Flash Attention 2 support are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "Sparse-layer construction in SwitchTransformers and a Python 3.13 IndentationError in DebertaV2Model are not the same bug." + }, + { + "left": "issue:44242", + "right": "issue:46897", + "accept": false, + "reason": "Both are training-loss related, but they concern different models and different loss-shaping mistakes." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "A missing import from image_utils and a Qwen3.5 chat crash are unrelated failures." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "Optional post_init handling in patchtsmixer and strict config rejection for granite_speech are different loading/configuration issues." + }, + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "A video-position-id regression and a Sam3LiteText memory overflow do not indicate the same underlying defect." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale warning in paged generate and a Whisper alignment/generation crash are different generate-path issues." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "Torch.compile SDPA failure in Bamba and local dataset loading failure are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "An A800 cudaErrorAssert under auto device mapping is not the same as a 2^24 category limit runtime error." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Sam3PixelDecoder using a dummy Conv2d and OmDet-Turbo processor/model input-size mismatch are different bugs." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "A video-input regression across Qwen VL models and a serve KV-cache indexing crash are separate code paths." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "SwitchTransformers sparse-layer config and local dataset loading in the training script are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Flash-attention auxiliary-state handling and xLSTM return-value unpacking are different runtime errors." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Both are memory-related and both involve Sam3 variants, but they are different models with different triggers and likely different fixes." + }, + { + "left": "issue:45381", + "right": "issue:46655", + "accept": false, + "reason": "These are both video-processor/model alignment issues, but the affected models and padding/position-id mechanics differ." + }, + { + "left": "issue:43854", + "right": "issue:44792", + "accept": false, + "reason": "A unit-test model-loading problem and a failed image-generation test for janus are not the same underlying bug." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "A SigLIP2 docs/model-processor mismatch and a local dataset loading failure in a training script are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 39, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29870", + "issue:32090", + "issue:35141", + "issue:36246", + "issue:38617", + "issue:41628", + "issue:42175", + "issue:42796", + "issue:43295", + "issue:43404", + "issue:43526", + "issue:43540", + "issue:43643", + "issue:43673", + "issue:43854", + "issue:44112", + "issue:44246", + "issue:44279", + "issue:44462", + "issue:44464", + "issue:44496", + "issue:44521", + "issue:44734", + "issue:44737", + "issue:44779", + "issue:44938", + "issue:44964", + "issue:45003", + "issue:45127", + "issue:45216", + "issue:45276", + "issue:45290", + "issue:45357", + "issue:45381", + "issue:45542", + "issue:45563", + "issue:45584", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:45907", + "issue:46234", + "issue:46489", + "issue:46697", + "issue:46752", + "issue:46858" + ], + "soft_pairs": [ + "issue:45127|issue:45357", + "issue:44464|issue:45563", + "issue:43295|issue:45276", + "issue:44779|issue:46489", + "issue:43540|issue:44734", + "issue:44521|issue:45381", + "issue:45290|issue:46752", + "issue:44112|issue:45600", + "issue:45276|issue:45584", + "issue:43526|issue:46697", + "issue:25251|issue:29870", + "issue:44462|issue:45907", + "issue:44737|issue:45803", + "issue:44938|issue:45542", + "issue:43673|issue:46858", + "issue:43643|issue:44496", + "issue:43854|issue:44964", + "issue:44246|issue:45003", + "issue:44964|issue:45600", + "issue:27957|issue:35141", + "issue:45632|issue:46234", + "issue:42796|issue:44521", + "issue:44112|issue:44964", + "issue:42175|issue:44279", + "issue:36246|issue:41628", + "issue:32090|issue:35141", + "issue:43404|issue:45216", + "issue:36246|issue:38617" + ], + "prepared_review_unit_hash": "5f6ca4d355cb5ba3f8c3b9c568201a154fa6e825a7d171cd49e77790466706c6", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36521, + "estimated_input_tokens": 9131, + "estimated_eval_tokens": 18518 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a grab bag of unrelated Transformers issues across tokenization, generation, multimodal processing, imports, and CI/dependency problems. None of the soft-edge pairs look like true duplicates, so all are rejected. The most representative issue is the open multimodal `apply_chat_template` assistant-mask bug.", + "confidence": 0.74, + "canonical_issue_reason": "issue:44521 is the most central representative: it is open, has the most inbound references among the set, and describes a concrete multimodal chat-template bug with a clear repro surface.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the strongest triage candidate overall because it is active, well-discussed, and a specific user-facing bug in a core multimodal API.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "Related around Qwen3.5 saving/weights, but these are different bugs: LoRA merge with extended vocab vs incorrect visual encoder keys in `save_pretrained`." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Both touch generation, but one is inconsistent compiled chunked generation outputs and the other is a stale warning about `num_return_sequences`." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Both mention model save/resize regressions, but they affect different code paths and different model families." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": false, + "reason": "Same tokenizer family, but one is a general DeepSeek v5 regression and the other is a DeepSeek-Coder v1-specific gap in a prior fix." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Different failures: Qwen3OmniMoe video-input processing vs `transformers serve` KV-cache continuation tensor indexing." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal-related, but one is chat-template assistant masks and the other is wrong `vision_position_ids` for Qwen2.5-VL video inputs." + }, + { + "left": "issue:45290", + "right": "issue:46752", + "accept": false, + "reason": "Both are `apply_chat_template` crashes, but the triggering inputs differ: tool-call assistant messages vs an empty list." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Unrelated CI failures: a stale GraniteSpeech device override test vs broken `auto_mappings.py` references to removed configs." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Different bugs: embedding resize not propagating vs Whisper generation failing on empty transcription after token alignment." + }, + { + "left": "issue:43526", + "right": "issue:46697", + "accept": false, + "reason": "Different image-processing bugs in different models: label reduction vs over-padding in Swin2SR." + }, + { + "left": "issue:25251", + "right": "issue:29870", + "accept": false, + "reason": "Distinct generation bugs: pipeline `top_k` nesting change vs Bart `decoder_attention_mask` initialization error." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "Different APIs and failures: `AutoTokenizer` ignoring `tokenizer.json` vs `list_repo_files` not forwarding `revision`." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "Different root causes: XLNet device placement bug vs a bare `except:` in Fuyu tensor conversion." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Different environment/dependency issues: Python 3.14 import failure vs TensorBoard installed without TensorFlow backend." + }, + { + "left": "issue:43673", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache handling, but one is missing generation cache in v5.0.0 and the other is a device-tensor slice arithmetic bug in `_prefill`." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Different config-loading failures: missing fields with `trust_remote_code=True` vs unrecognized model type in config.json." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Both are model-load failures, but they involve different models and different failure modes." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Different import-related problems: slow `import transformers` vs unsafe `sys.modules` access in `modeling_utils`." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "A model loading regression and a CI break from removed auto-mapping references are not the same bug." + }, + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "Unrelated model/runtime issues: XLMRoberta Flash Attention 2 support vs embedding reinitialization after resize when untied." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "Different issues in different layers: trust-remote-code cache path collision vs missing `is_falcon_mamba_ssm_available` export." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer BPE-marker handling and multimodal chat-template assistant masks are separate bugs." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flake in GraniteSpeech is unrelated to the Phi-4-multimodal model load failure." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "One is a packaging/backend install issue and the other is a vague dependency problem; not the same concrete bug." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and different modules." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Different subsystems entirely: trainer GPU broadcast of `None` vs tied/untied embedding resize reinitialization." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Different regression targets: Mistral3 weight tying vs Qwen3.5 checkpoint saving correctness." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and different modules." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 40, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:41950", + "issue:42907", + "issue:43295", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43526", + "issue:43531", + "issue:43582", + "issue:43650", + "issue:43673", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44220", + "issue:44279", + "issue:44295", + "issue:44327", + "issue:44336", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44792", + "issue:44933", + "issue:44987", + "issue:45042", + "issue:45083", + "issue:45335", + "issue:45375", + "issue:45446", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45600", + "issue:45759", + "issue:46082", + "issue:46121", + "issue:46123", + "issue:46369", + "issue:46424", + "issue:46506", + "issue:46515", + "issue:46693", + "issue:46752", + "issue:46932" + ], + "soft_pairs": [ + "issue:46082|issue:46121", + "issue:43531|issue:45600", + "issue:44295|issue:46506", + "issue:41950|issue:44016", + "issue:43673|issue:46424", + "issue:43295|issue:45335", + "issue:43582|issue:46932", + "issue:45083|issue:46506", + "issue:43976|issue:44987", + "issue:44016|issue:44220", + "issue:44496|issue:45042", + "issue:43673|issue:44336", + "issue:43526|issue:43901", + "issue:43901|issue:44496", + "issue:42907|issue:43441", + "issue:44683|issue:45446", + "issue:43650|issue:44220", + "issue:43381|issue:45593", + "issue:44496|issue:45759", + "issue:44792|issue:46424", + "issue:44016|issue:44062", + "issue:43441|issue:45588", + "issue:43976|issue:44279", + "issue:44521|issue:46693", + "issue:43824|issue:44933", + "issue:44279|issue:46515", + "issue:43479|issue:45375", + "issue:46123|issue:46369", + "issue:45335|issue:45584", + "issue:44327|issue:46752" + ], + "prepared_review_unit_hash": "eda2d5cd5e4efe0877baf226d0e8ee8118d344aae65f15953a353341d19775c4", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35950, + "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 soft pairs share broad Transformer/modeling themes but usually point to different subsystems, different failure modes, or different versions. None look like the same underlying bug/change, so I reject all soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "No strong canonical issue exists here; the items are not a single duplicate set, and the apparent similarities are too broad to justify merging them into one representative issue.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a good global representative because the cluster spans unrelated regressions and feature bugs across different components; any choice would be arbitrary.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Llama head_dim validation and rope-parameter JSON list handling are unrelated config-path bugs." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto_mappings CI breakage concern different models and code paths." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "A position_ids buffer access issue and CPU/CUDA output divergence are different problems with no clear shared bug." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification processor lookup and a notebook syntax error are unrelated; one is runtime logic, the other docs/notebook content." + }, + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Missing GenerationMixin cache in chunked_prefill is not the same as max_cache_length being honored only for quantized cache." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "These both involve multimodal/composite configs, but one is a processor/tokenizer regression and the other is a resize_token_embeddings propagation bug." + }, + { + "left": "issue:43582", + "right": "issue:46932", + "accept": false, + "reason": "Apple Silicon TypeError in caching_allocator_warmup is unrelated to Gemma4 quantized vision/audio dtype casting." + }, + { + "left": "issue:45083", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3 Omni MoE feature-length calculation and PPDocLayoutV3 CPU/CUDA mismatch are separate failures." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Python version incompatibility for Transformers 5.1.0 is not the same as loading a specific model failing on newer Transformers." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "A notebook syntax error and _torch_extract_fbank_features() failure are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Unrecognized model config loading and PIL backend processors requiring torchvision are different model-loading/backend issues." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "chunked_prefill cache behavior and ANSI escape generation in loading_report are unrelated code paths." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug and TextClassificationPipeline docs mismatch are not the same underlying issue." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Docs wording for return_all_scores and model-config recognition failure are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Saving dequantized Ministral/Devstral models is not the same as FlashAttention failures for Ministral-3." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and a wrong torch version check for AuxRequest import are distinct flex_attention issues." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "A vague data request issue does not match a concrete audio feature extraction bug." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary losses when denoising is off are unrelated training behaviors." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Unrecognized model config loading and AutoModelForCausalLM not unwrapping text_config are different composite-config problems." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "A failed janus image-generation test and cache max length semantics are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken special-argument duplication are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Ministral-3 FlashAttention failure and sink-less-model s_aux=None crash are both attention-related but involve different concrete failure paths." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python version support and a generic dependency issue do not identify the same bug." + }, + { + "left": "issue:44521", + "right": "issue:46693", + "accept": false, + "reason": "all-zero assistant masks in multimodal chat templating and FA2 long-context performance regression are different generation issues." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Importing Qwen2_5_VLForConditionalGeneration and a nonexistent image_utils import are separate missing-export problems." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "Generic dependency trouble and torch 2.4.1 incompatibility are not the same concrete issue." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Default vision/audio config initialization and strict-dropping of deepstack_visual_indexes are different config-field bugs." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "Final partial batch handling in MaskGenerationPipeline is unrelated to memory overflow in sam3-litetext." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "decoder.embed_tokens resize propagation and empty-transcription Whisper generation failure are different token/decoding bugs." + }, + { + "left": "issue:44327", + "right": "issue:46752", + "accept": false, + "reason": "QA top-k bounds error in decode_spans and empty-list apply_chat_template IndexError are unrelated edge cases." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 41, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:33290", + "issue:39692", + "issue:41950", + "issue:42796", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43577", + "issue:43650", + "issue:43653", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44075", + "issue:44242", + "issue:44279", + "issue:44291", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44466", + "issue:44683", + "issue:44779", + "issue:45020", + "issue:45081", + "issue:45200", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45357", + "issue:45405", + "issue:45542", + "issue:45588", + "issue:45593", + "issue:45656", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45907", + "issue:46082", + "issue:46209", + "issue:46399", + "issue:46512", + "issue:47000" + ], + "soft_pairs": [ + "issue:43299|issue:45216", + "issue:44279|issue:45405", + "issue:43295|issue:44779", + "issue:42796|issue:43650", + "issue:39692|issue:43901", + "issue:46082|issue:46399", + "issue:44075|issue:44368", + "issue:43577|issue:45216", + "issue:44351|issue:45278", + "issue:41950|issue:45020", + "issue:43976|issue:45405", + "issue:44779|issue:45812", + "issue:46082|issue:46512", + "issue:44466|issue:45276", + "issue:43299|issue:45357", + "issue:43381|issue:47000", + "issue:43650|issue:46209", + "issue:43653|issue:45593", + "issue:44242|issue:45200", + "issue:41950|issue:45823", + "issue:43299|issue:43854", + "issue:44462|issue:45081", + "issue:44016|issue:44291", + "issue:44683|issue:45542", + "issue:45588|issue:47000", + "issue:43479|issue:43525", + "issue:33290|issue:45656", + "issue:45081|issue:45907", + "issue:43441|issue:45834" + ], + "prepared_review_unit_hash": "5c4b69bf7d1f9060a4772a992d9b94edcdc506ae2fae03f55a111af56949524c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36480, + "estimated_input_tokens": 9120, + "estimated_eval_tokens": 18496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is highly heterogeneous: the candidate pairs mostly share only broad subsystem labels (tokenizers, configs, deepspeed, docs, multimodal models) but not the same concrete bug or change. I rejected all soft edges and found no clear canonical issue/PR representative for the whole 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:43299", + "right": "issue:45216", + "accept": false, + "reason": "Different models and failure modes: Qwen3VL MoE loading vs Qwen3.5 save_pretrained visual-encoder key serialization." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "One is a vague dependency issue; the other is a specific PEFT version bump/release mismatch. Not the same bug." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Both are regressions, but one is processor/tokenizer API breakage and the other is Deepseek tokenization output changes. Different code paths." + }, + { + "left": "issue:42796", + "right": "issue:43650", + "accept": false, + "reason": "Completely unrelated: tokenizer BPE marker handling vs a non-informative 'ADD THE DATA' issue." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Both mention docs, but one is SigLIP2 example errors and the other is a stale TextClassificationPipeline doc note. Different fixes." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "Different model/config bugs: LlamaConfig head_dim validation vs Qwen3.5 kernel registration failure." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Unrelated optimizer-argument handling vs a tie_word_embeddings warning on Qwen3.5 LoRA fine-tuning." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Different areas: Blip2 dtype propagation on load vs Qwen3.5 save_pretrained key layout." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but one is a missing HybridCache symbol and the other is broad import failures after upgrade." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Video-classification processor selection bug vs a broad remote_code regression; not the same underlying issue." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python compatibility failure vs a dependency version bump. No shared concrete bug." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Different tokenizer regressions affecting different model families (Deepseek vs Granite)." + }, + { + "left": "issue:46082", + "right": "issue:46512", + "accept": false, + "reason": "Head_dim validation is unrelated to from_config(dtype=...) leaving config.dtype stale." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Serialization of lm_head.weight by device is unrelated to resize_token_embeddings not updating Gemma4 embeddings." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both touch Qwen3.5/Qwen models, but one is model loading and the other is save_pretrained visual-encoder key output. Different bugs." + }, + { + "left": "issue:43381", + "right": "issue:47000", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction is unrelated to RTDetrLoss infeasible cost-matrix failure." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Not related: placeholder issue vs FP8 inference failure on B200." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "BigBirdTokenizer special-token decode bug is unrelated to D-FINE auxiliary-loss omission." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Both concern training losses, but one is router load-balancing loss gating and the other is mm_token_type_ids defaulting in Gemma4." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Different multimodal model bugs: processor lookup in video pipeline vs Gemma4 PLE device placement." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different models and symptoms: Qwen3VL MoE loading vs GLM-4.7-Flash unit-test loading." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "Both tokenizer-related, but AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashing are separate code paths." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error is unrelated to init_empty_weights/_is_hf_initialized TypeError." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 is unrelated to a TensorBoard/TensorFlow backend import issue." + }, + { + "left": "issue:45588", + "right": "issue:47000", + "accept": false, + "reason": "Flash-attention s_aux=None crash is a different failure mode than RTDetrLoss infeasible cost matrix." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Both are config-related, but Phi4Multimodal default sub-config initialization and Llama4Config pad_token_id absence are distinct bugs." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "DeepSpeed-related optimizer OOM and optimizer.step being called twice are not the same concrete problem." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regex patch crash is unrelated to list_repo_files failing to forward revision." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "Ministral FlashAttention failure and Kosmos2.5 OCR long-input index error are separate model-specific bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 42, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:42175", + "issue:42796", + "issue:42994", + "issue:43317", + "issue:43441", + "issue:43525", + "issue:43643", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44162", + "issue:44186", + "issue:44188", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44297", + "issue:44336", + "issue:44373", + "issue:44496", + "issue:44589", + "issue:44623", + "issue:44671", + "issue:44749", + "issue:44779", + "issue:44987", + "issue:45083", + "issue:45103", + "issue:45325", + "issue:45362", + "issue:45372", + "issue:45397", + "issue:45542", + "issue:45600", + "issue:45712", + "issue:45823", + "issue:45859", + "issue:46291", + "issue:46399", + "issue:46655", + "issue:46726", + "issue:46851", + "issue:47000" + ], + "soft_pairs": [ + "issue:44336|issue:46851", + "issue:42994|issue:43317", + "issue:44623|issue:44749", + "issue:45859|issue:46851", + "issue:45103|issue:45372", + "issue:44261|issue:44263", + "issue:44623|issue:45397", + "issue:42175|issue:43976", + "issue:45600|issue:45712", + "issue:43901|issue:44246", + "issue:44671|issue:44779", + "issue:43673|issue:43761", + "issue:43525|issue:43881", + "issue:43723|issue:44162", + "issue:44016|issue:45823", + "issue:44188|issue:44336", + "issue:42796|issue:43441", + "issue:44186|issue:47000", + "issue:43824|issue:43976", + "issue:44987|issue:45542", + "issue:43525|issue:43643", + "issue:44336|issue:44373", + "issue:44589|issue:45362", + "issue:45325|issue:46851", + "issue:46655|issue:46726", + "issue:46291|issue:46399", + "issue:44246|issue:44496", + "issue:45083|issue:46851", + "issue:44297|issue:45397" + ], + "prepared_review_unit_hash": "f23620dd972223b62969ba0d6affdac68793eb37bcebf5d5915608558988f809", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36409, + "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 highly heterogeneous and does not contain any clear duplicate pairs. The soft matches mostly reflect superficial title similarity around transformers v5, tokenizer, config, or loading issues, but they are different bugs or changes.", + "confidence": 0.92, + "canonical_issue_reason": "If forced to choose one issue as the representative anchor, issue 44246 is the broadest user-facing regression and has the most discussion, but the overall set is not a true duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44246 is the best single issue to anchor triage because it is broadly relevant, generic, and well-discussed; however, it should not be treated as a duplicate representative for the other items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI escape-code rendering in loading_report is unrelated to Qwen3.5 variable-length position_ids." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Both involve quantization/offload, but one is about saving quantized models and the other about loading a dequantized model with device_map=auto." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "processor.save_pretrained file-missing behavior is unrelated to a slowdown after upgrading Transformers." + }, + { + "left": "issue:45859", + "right": "issue:46851", + "accept": false, + "reason": "Both are Qwen3.5-related, but CPU offload rotary embedding handling is a different code path from variable-length position_ids." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "auto_docstring future-annotations crashes are unrelated to Gemma 4 processor loading/import failures." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "These describe different model/helper bugs: rms_norm_eps precision drift vs torch.split return handling." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Missing processor.save_pretrained outputs is unrelated to Gemma-4 zero3 from_pretrained loading." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "TensorFlow backend packaging and Python-version compatibility are different issues." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Both touch repo maintenance, but removed auto_mappings entries and leftover dummy classes are distinct problems." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Documentation mismatch for return_all_scores is unrelated to intermittent import-time slowness." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "These are separate model regressions affecting different tokenization/model outputs." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Chunked-prefill cache handling and CLIP hidden_states propagation are different functional bugs." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "A missing pad_token_id attribute in Llama4Config is not the same as GLM-4V loading failure." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "AutoTokenizer loading regression and ESM2 model breakage are unrelated issues." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "Notebook syntax error and Gemma4 PLE device mismatch are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "torch.compile attention-kernel divergence has no connection to ANSI code output in loading_report." + }, + { + "left": "issue:42796", + "right": "issue:43441", + "accept": false, + "reason": "Tokenizer BPE marker handling and FlashAttention failure on Ministral-3 are different bugs." + }, + { + "left": "issue:44186", + "right": "issue:47000", + "accept": false, + "reason": "LayoutLMv2Tokenizer padding/truncation crashes are unrelated to RTDetrLoss infeasible cost matrix errors." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Missing Qwen2_5_VL export is unrelated to the Python 3.9/3.10 compatibility issue." + }, + { + "left": "issue:44987", + "right": "issue:45542", + "accept": false, + "reason": "Model loading regression and TensorFlow backend package detection are separate problems." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "A model config missing pad_token_id is not the same as AutoConfig losing fields under trust_remote_code." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI-code output in a utility and a wrong docstring are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage lookup failure and Qwen3.5 chat crash are different failure modes." + }, + { + "left": "issue:45325", + "right": "issue:46851", + "accept": false, + "reason": "Both concern position_ids, but Qwen2.5-VL temporal scaling and Qwen3.5 variable-length inputs are different bugs." + }, + { + "left": "issue:46655", + "right": "issue:46726", + "accept": false, + "reason": "Video padding calculation and missing vision backend/image processor errors are unrelated." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "kernels v0.15 dependency versioning and missing decorator registration are different kernel-related issues." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import latency and model config recognition failure are not the same underlying bug." + }, + { + "left": "issue:45083", + "right": "issue:46851", + "accept": false, + "reason": "These are separate Qwen-related position handling bugs in different model families and code paths." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to Gemma-4 zero3 loading." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 43, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43257", + "issue:43299", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43784", + "issue:43901", + "issue:43976", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44451", + "issue:44485", + "issue:44496", + "issue:44779", + "issue:44871", + "issue:45003", + "issue:45072", + "issue:45103", + "issue:45137", + "issue:45292", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45561", + "issue:45563", + "issue:45588", + "issue:45593", + "issue:46132", + "issue:46399", + "issue:46506", + "issue:46614", + "issue:46655", + "issue:46665", + "issue:46752", + "issue:46821", + "issue:46829", + "issue:46851" + ], + "soft_pairs": [ + "issue:41628|issue:43976", + "issue:43784|issue:45542", + "issue:46399|issue:46851", + "issue:45561|issue:45563", + "issue:43540|issue:46752", + "issue:43526|issue:45003", + "issue:45103|issue:46399", + "issue:45072|issue:46829", + "issue:46614|issue:46665", + "issue:44295|issue:45397", + "issue:43531|issue:45362", + "issue:45588|issue:46851", + "issue:44246|issue:44871", + "issue:42491|issue:43299", + "issue:44336|issue:44485", + "issue:45003|issue:46821", + "issue:44242|issue:45593", + "issue:42886|issue:46132", + "issue:43550|issue:46506", + "issue:44779|issue:45292", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:43901|issue:45003", + "issue:44451|issue:46752", + "issue:43257|issue:45137", + "issue:41950|issue:44220", + "issue:43540|issue:46655", + "issue:44496|issue:45003" + ], + "prepared_review_unit_hash": "6990eb3a6df6a735b67d43d406f2b4c3cce382e9e57901247c6ae5e65dc234e8", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35999, + "estimated_input_tokens": 9000, + "estimated_eval_tokens": 18256 + }, + "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; they span unrelated import errors, model-loading regressions, processor bugs, and docs/runtime issues.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 45561 is the strongest standalone representative: it is open, has clear repro context, multiple inbound references, and a concrete CI/testing-utils failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45561 is the best issue to keep as a canonical reference for future duplicates because it is specific, actionable, and appears broadly relevant to test infrastructure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "Different failures: a missing AutoImageProcessor export vs a Python-version compatibility issue; not the same code-path bug." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both are import-time errors, but they concern different missing backends/symbols in different integrations ('nn' vs undefined tf backend)." + }, + { + "left": "issue:46399", + "right": "issue:46851", + "accept": false, + "reason": "Both involve Qwen3.5, but one is kernel registration/use_kernels and the other is variable-length position_ids; distinct bugs." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "Test utility race on captured_info.txt is unrelated to a stale generation warning for num_return_sequences." + }, + { + "left": "issue:43540", + "right": "issue:46752", + "accept": false, + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to apply_chat_template crashing on an empty list." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and unsafe sys.modules access in modeling_utils are different subsystems and failures." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "Auto-docstring annotation handling is unrelated to Qwen3.5 rotary kernel registration." + }, + { + "left": "issue:45072", + "right": "issue:46829", + "accept": false, + "reason": "Different model families and symptoms: dtype mismatch in bfloat16 inference vs hidden_states output wiring." + }, + { + "left": "issue:46614", + "right": "issue:46665", + "accept": false, + "reason": "A stale phonemizer backend cache bug is not the same as inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer reading error and gemma-4 ZeRO3 from_pretrained failure are distinct." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Qwen3-MoE sliding_window handling and Qwen3.5 chat crash are different code paths." + }, + { + "left": "issue:45588", + "right": "issue:46851", + "accept": false, + "reason": "Flash-attention s_aux=None crash is unrelated to variable-length position_ids handling." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import latency and eos_token_id config mismatch are unrelated issues." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both mention Qwen3 MoE loading, but one is a LoRA/main-branch compatibility problem and the other is a Qwen3VL-MoE regression; too different to merge." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI-code cleanup in loading_report is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:45003", + "right": "issue:46821", + "accept": false, + "reason": "Unsafe sys.modules access and TimesFM window_size AttributeError are unrelated." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Both concern missing auxiliary losses, but for different models and different conditions; not the same concrete bug." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline tokenizer cache loading and AttentionInterface.register behavior change are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to CPU-vs-CUDA output differences in PPDocLayoutV3." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Deepseek tokenizer regression and resize_token_embeddings not updating output_embeddings are different bugs." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Missing Qwen2.5 VL image processor export and missing is_offline_mode export are separate import errors." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Both are missing-name import errors, but they involve different exported symbols and modules." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mismatch for return_all_scores and sys.modules access are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:46752", + "accept": false, + "reason": "Model-loading failure for ScandiBERT is unrelated to apply_chat_template on an empty list." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed and a ZeRO3 deque IndexError are different DeepSpeed failures." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification looking for image processors and a torch_extract_fbank_features issue are different pipeline bugs." + }, + { + "left": "issue:43540", + "right": "issue:46655", + "accept": false, + "reason": "Both involve video processing, but they affect different models and different failure mechanisms (ValueError vs padding calculation)." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized model_type in config loading is unrelated to unsafe 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": 26, + "review_unit_index": 44, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:36246", + "issue:42199", + "issue:43262", + "issue:43295", + "issue:43352", + "issue:43475", + "issue:43526", + "issue:43596", + "issue:43653", + "issue:44038", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44623", + "issue:44749", + "issue:44857", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45250", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45459", + "issue:45561", + "issue:45600", + "issue:45759", + "issue:45859", + "issue:46369", + "issue:46399", + "issue:46506", + "issue:46655", + "issue:46726", + "issue:46851", + "issue:46894", + "issue:46897", + "issue:47000" + ], + "soft_pairs": [ + "issue:45375|issue:45759", + "issue:43262|issue:44246", + "issue:44479|issue:46655", + "issue:45003|issue:45561", + "issue:44162|issue:45200", + "issue:44295|issue:44749", + "issue:45859|issue:46399", + "issue:45292|issue:47000", + "issue:45137|issue:45600", + "issue:43475|issue:46369", + "issue:43295|issue:45292", + "issue:36246|issue:43596", + "issue:44466|issue:45335", + "issue:44623|issue:46726", + "issue:44877|issue:45375", + "issue:44242|issue:44857", + "issue:43526|issue:46897", + "issue:44038|issue:46506", + "issue:44987|issue:45459", + "issue:29870|issue:30333", + "issue:42199|issue:46851", + "issue:43352|issue:45250", + "issue:43526|issue:46894", + "issue:44060|issue:44077", + "issue:44368|issue:44977", + "issue:43653|issue:44162" + ], + "prepared_review_unit_hash": "6313c86199f85de96c9986887615b866d761b883760265b4df737cd119c4456a", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35794, + "estimated_input_tokens": 8949, + "estimated_eval_tokens": 18154 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The proposed soft links are all false positives: the issues span unrelated models, processors, training bugs, and infrastructure problems with only superficial wording overlap.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the candidates are heterogeneous and the soft links are weak or superficial.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45561 is the most active and well-specified standalone bug in the set, but it does not serve as a duplicate anchor for the rest.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Both involve Qwen multimodal config/model loading, but they hit different bugs: strict-config field dropping vs unwrapping text_config in from_config." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio chat-template sampling-rate handling and import-time slowdown are unrelated issues." + }, + { + "left": "issue:44479", + "right": "issue:46655", + "accept": false, + "reason": "Both are video regressions, but for different models and different failure causes (general Qwen video regression vs SmolVLM padding calculation)." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "sys.modules access safety and a pytest-xdist file race are distinct subsystems and bugs." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma 4 mm_token_type_ids defaults are unrelated model-specific issues." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "A position_ids buffer access error and a post-upgrade performance regression are not the same bug." + }, + { + "left": "issue:45859", + "right": "issue:46399", + "accept": false, + "reason": "Both are Qwen3.5-related, but CPU offload rotary embedding support and apply_rotary_pos_emb/kernel registration are different code paths." + }, + { + "left": "issue:45292", + "right": "issue:47000", + "accept": false, + "reason": "Embedding-resize propagation and RTDetr loss infeasible-cost-matrix crashes are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque underflow and stale auto_mappings config references are different problems." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "Missing fpn_position_embeddings and memory overflow in sam3-litetext are different SAM3 issues." + }, + { + "left": "issue:43295", + "right": "issue:45292", + "accept": false, + "reason": "Custom processor/tokenizer regression and output embedding resize propagation are different APIs and failure modes." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Qwen2.5-VL import failure and BertModel ZeRO3 initialization IndexError are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both mention embeddings/tied weights, but one is serialization inconsistency and the other is resize_token_embeddings not updating decoder embeddings." + }, + { + "left": "issue:44623", + "right": "issue:46726", + "accept": false, + "reason": "Missing files on save_pretrained and missing vision backend handling in AutoProcessor are different bugs." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Strict config loading for granite_speech and a dropped Qwen config field are separate config-handling issues." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Load-balancing loss omission and an AMP/CUDA crash in LwDetrImageLoss are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:46897", + "accept": false, + "reason": "Beit label reduction and Florence2 training-loss double-shift are different preprocessing/training bugs." + }, + { + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "A transformers/Qwen3-VL-Moe bug and a CPU/CUDA output discrepancy in PPDocLayoutV3 are not the same issue." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model loading failure on transformers>=5.1.0 and tokenizer error masking when protobuf is missing are different failure modes." + }, + { + "left": "issue:29870", + "right": "issue:30333", + "accept": false, + "reason": "Bart generate() UnboundLocalError and MLflow job-status reporting are unrelated." + }, + { + "left": "issue:42199", + "right": "issue:46851", + "accept": false, + "reason": "DETR cardinality logic and Qwen3.5 variable-length position_ids are different model bugs." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Unsupported Flash Attention 2.0 for Nemotron and a generic Flash Attention topic are not the same concrete bug." + }, + { + "left": "issue:43526", + "right": "issue:46894", + "accept": false, + "reason": "Beit label reduction and CohereASR training-loss double-shift are distinct issues." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Incorrect tied-weights warning and patchtsmixer post_init policy are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "A tie_word_embeddings warning in LoRA fine-tuning and a flash-attention generation regression are different problems." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBird tokenizer special-token decoding and ESM2 breakage are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 45, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:42886", + "issue:43065", + "issue:43408", + "issue:43475", + "issue:43540", + "issue:43646", + "issue:43701", + "issue:43723", + "issue:43906", + "issue:43931", + "issue:43994", + "issue:44162", + "issue:44188", + "issue:44297", + "issue:44361", + "issue:44368", + "issue:44479", + "issue:44483", + "issue:44623", + "issue:44749", + "issue:44871", + "issue:44877", + "issue:44995", + "issue:45103", + "issue:45200", + "issue:45229", + "issue:45397", + "issue:45412", + "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" + ], + "soft_pairs": [ + "issue:44479|issue:46170", + "issue:45412|issue:46123", + "issue:45397|issue:45656", + "issue:42886|issue:45229", + "issue:43065|issue:43408", + "issue:43475|issue:43540", + "issue:44877|issue:46170", + "issue:43723|issue:46710", + "issue:46170|issue:46832", + "issue:43931|issue:44368", + "issue:44188|issue:46202", + "issue:44368|issue:45563", + "issue:44995|issue:45200", + "issue:44483|issue:46202", + "issue:43906|issue:46202", + "issue:44188|issue:46421", + "issue:45987|issue:46082", + "issue:44995|issue:45588", + "issue:45593|issue:47068", + "issue:45685|issue:46431", + "issue:45103|issue:46858", + "issue:44361|issue:44749", + "issue:43065|issue:46123", + "issue:44297|issue:44623", + "issue:43994|issue:46726", + "issue:44749|issue:45656", + "issue:43701|issue:44749", + "issue:46752|issue:47020", + "issue:43646|issue:44162", + "issue:44623|issue:45656", + "issue:44871|issue:46132" + ], + "prepared_review_unit_hash": "04d2cf586e4c8f81e12c57d24455d22451a46c8fa0abed39abd8cf05ef27fd70", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37159, + "estimated_input_tokens": 9290, + "estimated_eval_tokens": 18836 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous, with one small tokenizer/processor serialization subgroup and many unrelated issues (vision, cache, deepspeed, Gemma, RT-DETR, etc.). No soft edges look like the same underlying bug/change, so all candidate merges are rejected. Issue #44297 is the best representative of the tokenizer serialization subgroup.", + "confidence": 0.34, + "canonical_issue_reason": "#44297 is the clearest core bug in the tokenizer save/load path: it directly describes incorrect tokenizer metadata being written, which can plausibly underlie nearby tokenizer serialization/loading complaints.", + "canonical_pr_reason": null, + "best_issue_reason": "#44297 best captures the most coherent subcluster around tokenizer/processor persistence and loading; it is concrete, actionable, and closest to the shared failure mode among the few related items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Different subsystems and failure modes: video-input regression in Qwen VL models vs Windows/ROCm import/type errors." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "Memory reclamation in RT-DETR is unrelated to MaskGenerationPipeline's partial-batch `is_last` logic." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma-4 Zero3 loading and deepspeed optimizer double-step are different bugs in different code paths." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline tokenizer cache loading is unrelated to Gemma4 multi-GPU CUDA OOM." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "One is a dummy Conv2d implementation detail; the other is a model-type warning during loading." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "SAM3 vision encoder attribute error and Qwen3OmniMoe video-processing ValueError are separate model-specific failures." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading for granite_speech is unrelated to the Windows ROCm distributed import issue." + }, + { + "left": "issue:43723", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer loading failure in v5 and an accuracy regression after upgrade are not the same concrete bug." + }, + { + "left": "issue:46170", + "right": "issue:46832", + "accept": false, + "reason": "Platform import/dependency failure and RT-DETR feature-level mismatch are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-VL weight shape mismatch and a tie_word_embeddings warning in Qwen3.5 are different issues." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "torch.compile attention-kernel branching divergence is unrelated to Gemma pipeline roundtrip tokenization failure." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A configuration warning about tied embeddings does not match a stale generate warning for num_return_sequences." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "GlmMoeDsa cache staleness on second forward is unrelated to Gemma4 mm_token_type_ids defaults." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "Chat/completions API request rejection is unrelated to Gemma pipeline text roundtrip behavior." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "An isolated reproduction of another issue and a Gemma roundtrip bug are not the same underlying defect." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "Attention kernel branching divergence is unrelated to cache lazy initialization causing recompilation." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "StaticCache return-type bug and LlamaConfig head_dim validation are different code paths with different fixes." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "GlmMoeDsa stale indexer cache and flash_attention s_aux=None AttributeError are unrelated failures." + }, + { + "left": "issue:45593", + "right": "issue:47068", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior and a deepspeed host-synchronization concern are not the same bug." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "MPS histogram support in MoE is unrelated to apply_chat_template missing-template errors." + }, + { + "left": "issue:45103", + "right": "issue:46858", + "accept": false, + "reason": "Auto-docstring annotation handling and cache.get_seq_length device-tensor arithmetic are separate issues." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "MLukeTokenizer AttributeError on tasks and a performance regression after upgrade do not share a concrete root cause." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "SAM3 dummy Conv2d and MaskGenerationPipeline partial-batch dropping results are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "These are the closest pair, but one is tokenizer metadata mismatch while the other is processor save_pretrained missing files; related area, not the same bug." + }, + { + "left": "issue:43994", + "right": "issue:46726", + "accept": false, + "reason": "SigLIP2 nonsensical outputs and AutoProcessor missing-vision-backend handling are distinct model-loading/processing issues." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "A broad slowdown regression after upgrade is not the same as deepspeed calling optimizer.step twice." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "resume_from_checkpoint key mismatch and a training slowdown regression are unrelated." + }, + { + "left": "issue:46752", + "right": "issue:47020", + "accept": false, + "reason": "Empty-list chat-template crash and XLMRobertaTokenizer/Unigram construction error are different tokenizer bugs." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Custom model initialization breakage in Transformers 5 and ESM2 being broken are too broad and not the same defect." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Processor save_pretrained missing files has nothing to do with deepspeed optimizer stepping twice." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and AttentionInterface.register behavior changes are separate API/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": 26, + "review_unit_index": 46, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43873", + "issue:43986", + "issue:44038", + "issue:44188", + "issue:44351", + "issue:44373", + "issue:44623", + "issue:44661", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:44964", + "issue:45092", + "issue:45137", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45397", + "issue:45406", + "issue:45478", + "issue:45561", + "issue:45753", + "issue:46202", + "issue:46326", + "issue:46399", + "issue:46463", + "issue:46726", + "issue:46821", + "issue:46851", + "issue:46877", + "issue:47000", + "issue:47059" + ], + "soft_pairs": [ + "issue:45375|issue:45478", + "issue:42994|issue:43482", + "issue:44964|issue:45137", + "issue:43475|issue:43643", + "issue:44188|issue:46851", + "issue:43475|issue:46821", + "issue:43232|issue:43673", + "issue:39692|issue:46726", + "issue:28282|issue:30990", + "issue:43986|issue:46726", + "issue:45230|issue:45397", + "issue:45310|issue:45375", + "issue:44373|issue:46851", + "issue:44038|issue:44623", + "issue:46877|issue:47000", + "issue:41628|issue:44933", + "issue:42994|issue:43525", + "issue:46463|issue:46877", + "issue:44869|issue:45356", + "issue:43723|issue:46326", + "issue:42371|issue:45092", + "issue:43873|issue:44945", + "issue:45561|issue:47059", + "issue:43784|issue:46399", + "issue:41628|issue:44351", + "issue:44945|issue:45753", + "issue:45406|issue:46202", + "issue:45137|issue:47000", + "issue:36010|issue:41628", + "issue:44661|issue:47059", + "issue:44038|issue:45310" + ], + "prepared_review_unit_hash": "bc5b47989b019a6833fe3bd4a5790fb30c1a5ee1a1727c8abe3a038bb7f2a2f0", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36639, + "estimated_input_tokens": 9160, + "estimated_eval_tokens": 18576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose collection of unrelated Transformers bugs, with one clear duplicate family around Qwen3.5-Moe / Qwen3.5-MoeVision from_pretrained failures. The rest span different models, APIs, and subsystems, so they should not be merged together.", + "confidence": 0.77, + "canonical_issue_reason": "Issue 45310 is the best representative because it is the clearest, most general description of the recurring Qwen3.5-Moe loading regression (`transformers>=5.4.0` + `from_pretrained` failure), which matches the closest duplicate set in this cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45310 is the strongest global representative: it names the concrete model family, the affected version range, and the failing load path, making it the most useful anchor for the duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45375", + "right": "issue:45478", + "accept": true, + "reason": "Both reports point to the same Qwen3.5-Moe/Vision `from_pretrained` regression, with the missing `deepstack_visual_indexes` field being a concrete manifestation of the loading failure." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "These are different bugs: one is quantized model saving, the other is GGUF loading under Transformers v5." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Different failure modes and subsystems: multimodal model loading versus DeepSpeed ZeRO3 deque handling." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "One is a SAM 3 Vision encoder attribute error; the other is a generic `trust_remote_code` config field-loss issue. Similar symptom class, but not the same bug." + }, + { + "left": "issue:44188", + "right": "issue:46851", + "accept": false, + "reason": "Both involve generation/attention behavior, but they are distinct code paths and symptoms: compile-time kernel branching versus variable-length position IDs." + }, + { + "left": "issue:43475", + "right": "issue:46821", + "accept": false, + "reason": "Different models and different missing attributes (`fpn_position_embeddings` vs `window_size`)." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both touch generation internals, but one is an `_update_model_kwargs_for_generation` ordering bug and the other is a missing `GenerationMixin` cache issue in chunked prefill." + }, + { + "left": "issue:39692", + "right": "issue:46726", + "accept": false, + "reason": "The former is a SigLIP2 documentation/example problem; the latter is a runtime AutoProcessor backend-missing error." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "Missing PyTorch in the environment is not the same as Sentence Transformers getting stuck while loading." + }, + { + "left": "issue:43986", + "right": "issue:46726", + "accept": false, + "reason": "Both involve processor/backend dependencies, but the concrete issues differ: video-model AutoProcessor crash without torchvision versus missing-vision-backend image processor handling." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Too generic to merge: one is an empty bug report, the other is a specific Gemma-4 ZeRO3 from_pretrained failure." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": true, + "reason": "These are the same underlying Qwen3.5-Moe loading/config regression; 45375 provides a specific missing-field symptom for the broader from_pretrained failure described in 45310." + }, + { + "left": "issue:44373", + "right": "issue:46851", + "accept": false, + "reason": "A docstring correction is unrelated to a variable-length input runtime bug." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe loading failure and `processor.save_pretrained(...)` missing files are different problems." + }, + { + "left": "issue:46877", + "right": "issue:47000", + "accept": false, + "reason": "Import-time doc-lint output and RTDetrLoss infeasibility are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they concern different missing symbols in different modules (`AutoImageProcessor` vs image_utils export)." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized saving failure and missing `pad_token_id` on Llama4Config are unrelated." + }, + { + "left": "issue:46463", + "right": "issue:46877", + "accept": false, + "reason": "Incorrect decoder-input validation messaging is unrelated to import-time docstring diagnostics." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Whisper timestamp decoding crash and Kimi tokenizer codec regression are different tokenizer bugs." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer loading in v5 and `continue_final_message` behavior for mistral_common are separate issues." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API settings and InternVL2 meta-initialization incompatibility are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading behavior and pipeline-parallelism output corruption are different runtime issues." + }, + { + "left": "issue:45561", + "right": "issue:47059", + "accept": false, + "reason": "A pytest-xdist race in testing utils is unrelated to a CLI completion attribute error from huggingface_hub." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "Missing `nn` import in sentence-transformers and a Qwen3.5 kernel-registration decorator error are different bugs." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports: `AutoImageProcessor` versus `HybridCache`." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline parallelism output issues and Qwen3.5 infinite-loop behavior are not the same failure." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Gemma4 processor missing `_tokenizer` and Gemma pipeline roundtrip failure on a special token are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:47000", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque underflow and RTDetrLoss cost-matrix infeasibility are different error sources." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Different missing imports from different APIs (`GenerationMixin` vs `AutoImageProcessor`)." + }, + { + "left": "issue:44661", + "right": "issue:47059", + "accept": false, + "reason": "A model-adder failure in tokenizer mapping is unrelated to CLI completion breakage." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both are Qwen-family loading bugs, but the model/version and symptom details are different enough that they do not clearly describe the same underlying defect." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: the Qwen3.5-Moe/Qwen3.5-MoeVision from_pretrained issues form the only clearly related family, while the rest of the items are plausibly separate bugs across different models/subsystems. The accepted soft-edge links are conservative enough given the titles, especially the identical 45310/45478 reports and the nearby 45375 config-field regression." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 47, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:39692", + "issue:42175", + "issue:42673", + "issue:42831", + "issue:42898", + "issue:42994", + "issue:43232", + "issue:43479", + "issue:43704", + "issue:43723", + "issue:43824", + "issue:43976", + "issue:44016", + "issue:44079", + "issue:44222", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44743", + "issue:44779", + "issue:44792", + "issue:44857", + "issue:44869", + "issue:44933", + "issue:45081", + "issue:45103", + "issue:45310", + "issue:45405", + "issue:45561", + "issue:45593", + "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" + ], + "soft_pairs": [ + "issue:44857|issue:46877", + "issue:45081|issue:45103", + "issue:42994|issue:43824", + "issue:46202|issue:46701", + "issue:44222|issue:45310", + "issue:45593|issue:46877", + "issue:42673|issue:43704", + "issue:43232|issue:44792", + "issue:42831|issue:44222", + "issue:44496|issue:45561", + "issue:42175|issue:44933", + "issue:44743|issue:47068", + "issue:45561|issue:45600", + "issue:45405|issue:46877", + "issue:43976|issue:44933", + "issue:30333|issue:33290", + "issue:30990|issue:35141", + "issue:42898|issue:44568", + "issue:43479|issue:46829", + "issue:42175|issue:45712", + "issue:44623|issue:45310", + "issue:46077|issue:46123", + "issue:45600|issue:47059", + "issue:42673|issue:46693", + "issue:43723|issue:44568", + "issue:43479|issue:44869", + "issue:42898|issue:45812", + "issue:44016|issue:46231", + "issue:39692|issue:46909", + "issue:43479|issue:44079", + "issue:42898|issue:44779" + ], + "prepared_review_unit_hash": "8b2eb19ebb7773be25579094f23827bca6e54fd5e8c66b67b5c2d71866b3a63a", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37091, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated Transformers issue reports; only one soft edge is an exact duplicate pair about Qwen3ForCausalLM VRAM leakage in multiple dataloader threads.", + "confidence": 0.96, + "canonical_issue_reason": "issue:42673 is the best canonical issue because it is the fuller, earlier report and exactly matches issue:43704 by title and bug description.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the strongest representative of the only true duplicate in this cluster; it has the more complete report and the duplicate later issue can be consolidated into it.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44857", + "right": "issue:46877", + "accept": false, + "reason": "Different bugs: AMP/CUDA crash in LwDetrImageLoss vs auto_docstring printing diagnostics at import time." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Both are AttributeError reports, but they affect different code paths: Mistral regex patching vs auto_docstring processing." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Unrelated failures: quantized save_pretrained vs importing Qwen2_5_VLForConditionalGeneration." + }, + { + "left": "issue:46202", + "right": "issue:46701", + "accept": false, + "reason": "Both mention Gemma, but one is a tokenizer roundtrip issue and the other is an image-text-to-text generation pipeline failure." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "Both are MoE/FP8 related, but one is FP8 save_pretrained and the other is a from_pretrained config/load error." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "Auxiliary-loss behavior in D-FINE is unrelated to auto_docstring import-time stdout noise." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact duplicate: same title and same VRAM leak in Qwen3ForCausalLM when used across multiple dataloader threads." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Different generation problems: sync_gpus kwargs handling vs janus image generation test failure." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "Both involve FP8, but one is an accuracy concern and the other is a save_pretrained bug in MoE." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Unrecognized model config error is unrelated to a pytest-xdist captured_info.txt race." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Backend packaging for TensorFlow vs a missing image_utils import are separate issues." + }, + { + "left": "issue:44743", + "right": "issue:47068", + "accept": false, + "reason": "Recurrent state reset with cache is not the same as a potential host sync in deepspeed_sp_compute_loss." + }, + { + "left": "issue:45561", + "right": "issue:45600", + "accept": false, + "reason": "Test file race in captured_info.txt is unrelated to stale auto_mappings references causing CI failures." + }, + { + "left": "issue:45405", + "right": "issue:46877", + "accept": false, + "reason": "Version pinning to unreleased PEFT is unrelated to auto_docstring import-time diagnostics." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version compatibility and missing import are different failures." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "Both are training/runtime issues, but MLFlow reporting on failure and DeepSpeed Adafactor OOM are distinct bugs." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Sentence Transformers load hangs and token-embedding reinit on resize are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but they describe different concrete behaviors and code paths." + }, + { + "left": "issue:43479", + "right": "issue:46829", + "accept": false, + "reason": "Config initialization with None is unrelated to the VideoPrism output field mismatch." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Package/backend inclusion and leftover dummy classes are separate maintenance issues." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Missing processor save_pretrained files is not the same bug as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Causal attention masks in Gemma4 and final-partial-batch handling in MaskGenerationPipeline are different code paths." + }, + { + "left": "issue:45600", + "right": "issue:47059", + "accept": false, + "reason": "Auto_mappings references to removed configs and a huggingface_hub CLI completion error are unrelated." + }, + { + "left": "issue:42673", + "right": "issue:46693", + "accept": false, + "reason": "VRAM leak in dataloader threads is not the same as the FA2 long-context performance regression." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer loading in v5 and add_special_tokens/BOS-EOS behavior are adjacent but not the same concrete bug." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Multimodal config default initialization is unrelated to Whisper timestamp decoding crashing on a trailing replacement character." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both concern tokenizer regressions, but one is clean_up_tokenization_spaces behavior and the other is Granite token ID generation." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Notebook syntax error and a docs mismatch on set_training parameter names are different documentation issues." + }, + { + "left": "issue:39692", + "right": "issue:46909", + "accept": false, + "reason": "Both are broken examples, but they involve different models and different failure modes." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Config defaulting and ModelOutput key assignment are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but there is no evidence they are the same underlying bug or code fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: the only clear duplicate pair is issue:42673 / issue:43704, and the other soft-edge rejections are conservative and consistent with the issue titles/descriptions." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 48, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28282", + "issue:30333", + "issue:34689", + "issue:36331", + "issue:41628", + "issue:42222", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42898", + "issue:43116", + "issue:43295", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:44060", + "issue:44075", + "issue:44190", + "issue:44222", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44448", + "issue:44485", + "issue:44623", + "issue:44734", + "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:46710", + "issue:46762" + ], + "soft_pairs": [ + "issue:43784|issue:43824", + "issue:27957|issue:30333", + "issue:30333|issue:34689", + "issue:42898|issue:43295", + "issue:42222|issue:45200", + "issue:43874|issue:45092", + "issue:44295|issue:45656", + "issue:44734|issue:45103", + "issue:42898|issue:45356", + "issue:30333|issue:36331", + "issue:22355|issue:28282", + "issue:44222|issue:44297", + "issue:44075|issue:46231", + "issue:45405|issue:45712", + "issue:43116|issue:44190", + "issue:44060|issue:45125", + "issue:42831|issue:44623", + "issue:44933|issue:45712", + "issue:44297|issue:45656", + "issue:44448|issue:46132", + "issue:42796|issue:43723", + "issue:45245|issue:46132", + "issue:43479|issue:43644", + "issue:44222|issue:44623", + "issue:44360|issue:46762", + "issue:46132|issue:46710", + "issue:44373|issue:44485", + "issue:41628|issue:42757", + "issue:44297|issue:45310" + ], + "prepared_review_unit_hash": "ff3cd84b9553fc01066f0623d82da26cef852c545e87aad9ae25ba4c9a4ccdbb", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35724, + "estimated_input_tokens": 8931, + "estimated_eval_tokens": 18118 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate pairs are overwhelmingly unrelated; I would reject all soft edges. If a representative issue must be chosen, the clearest standalone regression report is the v4.57.5 processor/tokenizer breakage issue.", + "confidence": 0.95, + "canonical_issue_reason": "No strong duplicate cluster emerges from these issues; they span unrelated import errors, model-specific regressions, trainer bugs, and docs. If one representative issue is needed, issue 43295 is the most actionable because it describes a concrete regression with a clear failure path and affected version.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43295 is the best global representative: it is a specific, reproducible regression in a recent release, with a well-defined code path (processor.tokenizer / image passing) and clear user impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Different import failures and different subsystems: sentence-transformers import vs missing Qwen2.5-VL class." + }, + { + "left": "issue:27957", + "right": "issue:30333", + "accept": false, + "reason": "Unrelated problems: Flash Attention behavior for XLM-Roberta vs MLflow reporting after failed training jobs." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Different concrete bugs: failed-job MLflow tracking vs model loading regression for Llama 3.2 Vision." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Both are v5 regressions, but one is tokenizer cleanup behavior and the other is processor/tokenizer access with images; not the same bug." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Different multimodal issues: vitpose model breakage vs Gemma 4 token-type defaults." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Both are multimodal/v5 compatibility issues, but one is a missing image processor method and the other is remote-code meta-init incompatibility." + }, + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Unrelated: position_ids buffer access error vs optimizer stepping twice under deepspeed." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Serving KV-cache tensor indexing bug vs auto-docstring crash from future annotations; different code paths." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenizers, but one is generic v5 cleanup behavior and the other is a Kimi-K2.5 codec regression; not the same underlying bug." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "Different training issues: MLflow status handling vs compute_loss signature incompatibility." + }, + { + "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; separate causes." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 MoE serialization and the other is tokenizer_class metadata mismatch." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer behavior bug vs documentation typo in set_training; not the same issue." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "Unrelated: unreleased PEFT version pin vs leftover dummy classes leaking into dir(transformers)." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Different example-script failures: multi-label classification returning empty results vs local dataset loading in image classification." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Same family (Qwen), but different concrete failures: false tied-weights warning vs missing tensor-parallel plan." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "Different bugs: FineGrainedFP8 accuracy regression vs processor.save_pretrained missing files." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Both are import/cleanup related, but one is an image_utils import failure and the other is dummy class leakage; not the same defect." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata bug vs optimizer step-count bug; unrelated." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Different concerns: Pegasus v4/v5 output change vs AttentionInterface.register behavior change." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Both are tokenizer-related v5 reports, but one is a specific BPE-marker handling bug and the other is a broader AutoTokenizer.from_pretrained loading issue." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Unrelated: category-cardinality runtime error vs attention registration behavior." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Different code paths: multimodal config defaulting vs non-persistent buffer initialization." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8 MoE serialization and the other is missing processor files." + }, + { + "left": "issue:44360", + "right": "issue:46762", + "accept": false, + "reason": "Different numerical/implementation issues: DSA indexer lacking ReLU vs Minimax MSA attention mismatch." + }, + { + "left": "issue:46132", + "right": "issue:46710", + "accept": false, + "reason": "Different bugs: registered attention function behavior vs DeepSeek-R1-Distill-Llama accuracy regression." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring mismatch for position_ids vs GLM-5 RoPE implementation; not the same underlying bug." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but one is a missing transformers symbol and the other is a huggingface_hub symbol mismatch." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save_pretrained metadata bug vs Qwen3.5 MoE from_pretrained regression; different operations and failure modes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 49, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:30990", + "issue:38175", + "issue:42831", + "issue:42907", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43475", + "issue:43504", + "issue:43723", + "issue:44038", + "issue:44190", + "issue:44222", + "issue:44263", + "issue:44279", + "issue:44295", + "issue:44360", + "issue:44373", + "issue:44483", + "issue:44484", + "issue:44623", + "issue:44829", + "issue:44964", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45310", + "issue:45357", + "issue:45381", + "issue:45406", + "issue:45834", + "issue:46002", + "issue:46050", + "issue:46082", + "issue:46209", + "issue:46305", + "issue:46328", + "issue:46369", + "issue:46402", + "issue:46424", + "issue:46431", + "issue:46614", + "issue:46655", + "issue:46909", + "issue:46932", + "issue:47059" + ], + "soft_pairs": [ + "issue:43504|issue:47059", + "issue:43723|issue:44977", + "issue:45071|issue:46082", + "issue:43122|issue:46209", + "issue:46614|issue:46932", + "issue:44222|issue:44484", + "issue:44263|issue:45092", + "issue:44190|issue:46082", + "issue:43232|issue:46424", + "issue:46209|issue:46431", + "issue:43232|issue:45834", + "issue:44295|issue:44623", + "issue:30990|issue:38175", + "issue:44263|issue:44373", + "issue:46369|issue:46909", + "issue:46305|issue:46655", + "issue:43475|issue:46305", + "issue:44263|issue:44360", + "issue:42907|issue:46050", + "issue:42831|issue:45310", + "issue:45092|issue:45357", + "issue:44279|issue:46328", + "issue:22355|issue:27957", + "issue:44483|issue:45406", + "issue:44964|issue:47059", + "issue:42831|issue:44038", + "issue:45381|issue:46305", + "issue:44829|issue:46002", + "issue:43408|issue:46402" + ], + "prepared_review_unit_hash": "d69cf30596f4f277c6d77cc5928f6fbd2df6f370a6e0e77300e30289c01618b4", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36022, + "estimated_input_tokens": 9006, + "estimated_eval_tokens": 18268 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate set is a loose collection of Transformers bug reports with only superficial thematic overlap; none of the soft pairs look like true duplicates of the same underlying bug/change, so all soft edges are rejected.", + "confidence": 0.92, + "canonical_issue_reason": "issue:46305 is the clearest standalone bug report here: it names a precise off-by-one in a specific masking function, has an explicit expected-vs-actual behavior, and is concrete enough to serve as the best issue representative among otherwise unrelated reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46305 is the strongest issue to keep as the representative artifact because it is specific, reproducible, and narrowly scoped to one code path, unlike the other candidates which are either broader, more model-specific, or less directly actionable.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43504", + "right": "issue:47059", + "accept": false, + "reason": "Different failures in unrelated areas: pretrained model loading for BEiT vs CLI completion support in Hugging Face Hub integration." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "One is tokenizer loading in v5, the other is Qwen3.5 generation with flash-attention; different model paths and symptoms." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "Both involve config validation in v5, but the concrete bugs differ: PretrainedConfig type checking vs explicit head_dim rejection in LlamaConfig." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenization regression and FP8 inference failure are unrelated bugs in different subsystems." + }, + { + "left": "issue:46614", + "right": "issue:46932", + "accept": false, + "reason": "Stale phonemizer language state is unrelated to Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save_pretrained failure is not the same as the max_shard_size default question." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "GlmMoeDsaIndexer split-return bug is different from remote-code checkpoint incompatibility during meta initialization." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Dataset loading in a training script is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation kwargs after sync_gpus and cache max length handling are different code paths with different behavior." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "FP8 inference on B200 and processor.save_pretrained missing files are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation model-kwarg handling is unrelated to Kosmos2.5 OCR length indexing." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Position_ids buffer access and processor.save_pretrained file omission are different bugs." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading stalls and siglip2 zero probabilities are different model/runtime issues." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "A DSA indexer return-value bug is not the same as a wrong docstring for position_ids." + }, + { + "left": "issue:46369", + "right": "issue:46909", + "accept": false, + "reason": "SAM3 memory overflow and a broken DIA documentation example are unrelated." + }, + { + "left": "issue:46305", + "right": "issue:46655", + "accept": false, + "reason": "Both touch video processing, but one is a sliding-window off-by-one and the other is a padding-calculation bug." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "SAM3 encoder attribute missing and MoonshineStreaming mask off-by-one are different model/components." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Same broad DSA indexer area, but one reports split-return semantics and the other a missing ReLU; not the same bug." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Both mention save/load of quantized models, but the affected mechanisms and failure modes differ." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy issues and Qwen3.5 from_pretrained errors are unrelated." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Meta-init incompatibility for old remote-code checkpoints is not the same as incorrect visual encoder keys saved by Qwen3.5." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A generic dependency issue is unrelated to SequenceFeatureExtractor.pad inefficiency." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Missing transformers.onnx module and XLMRoberta Flash Attention 2 behavior are unrelated issues." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat completions request rejection and Gemma4Processor missing _tokenizer are different server/processor failures." + }, + { + "left": "issue:44964", + "right": "issue:47059", + "accept": false, + "reason": "Phi-4 multimodal loading and CLI completion breakage are unrelated." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FP8 accuracy regression and Qwen3-VL-Moe v5 bug are different model-specific issues." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Both concern vision/video positions, but one is qwen2.5-vl vision_position_ids and the other is MoonshineStreaming masking; not the same code path." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Flash-attention-3 degenerate training and NaNs in classification heads are different training stability bugs." + }, + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "sam3_tracker/sam3_video model-type mismatch is unrelated to SAM3 text encoder weights not loading." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 50, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28218", + "issue:30990", + "issue:33357", + "issue:39401", + "issue:41628", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43232", + "issue:43329", + "issue:43408", + "issue:43540", + "issue:43576", + "issue:43644", + "issue:43646", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:43976", + "issue:44060", + "issue:44188", + "issue:44261", + "issue:44295", + "issue:44297", + "issue:44327", + "issue:44373", + "issue:44488", + "issue:44568", + "issue:44945", + "issue:44964", + "issue:45092", + "issue:45278", + "issue:45381", + "issue:45593", + "issue:45753", + "issue:46002", + "issue:46305", + "issue:46369", + "issue:46909", + "issue:47059" + ], + "soft_pairs": [ + "issue:41628|issue:43824", + "issue:44261|issue:45092", + "issue:43644|issue:46305", + "issue:42831|issue:45753", + "issue:44327|issue:47059", + "issue:22355|issue:28218", + "issue:39401|issue:44568", + "issue:42886|issue:42994", + "issue:44188|issue:44373", + "issue:42886|issue:43066", + "issue:42757|issue:45278", + "issue:43408|issue:46369", + "issue:27957|issue:30990", + "issue:42796|issue:43122", + "issue:45381|issue:46909", + "issue:43232|issue:45593", + "issue:43540|issue:46909", + "issue:43646|issue:46002", + "issue:42907|issue:44945", + "issue:30990|issue:33357", + "issue:44295|issue:44297", + "issue:43976|issue:44945", + "issue:43329|issue:47059", + "issue:42907|issue:44964", + "issue:43576|issue:43784", + "issue:43756|issue:44060", + "issue:39401|issue:44488", + "issue:41628|issue:45278" + ], + "prepared_review_unit_hash": "322cedd2cc06ce15b88b64252e7bbeef01660da489df270cf5d0127bf4a38cf3", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35153, + "estimated_input_tokens": 8789, + "estimated_eval_tokens": 17834 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are mostly unrelated regression reports that only look similar at the title level. None of the soft-similarity pairs are strong enough to merge as duplicates. If a single issue must be chosen as representative, issue 42886 is the clearest standalone bug report, but there is no true cluster-wide canonical.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 42886 is the most concrete, reproducible bug report in the set (version-specific offline cache failure), so it is the least ambiguous representative if one issue must be picked.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 42886 is the strongest standalone candidate: it has a precise failure mode, a pinned version, and an actionable reproduction path. The rest of the set is too heterogeneous to serve as a shared canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different symbols and different failure contexts; not the same bug." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "One is a q_a_layernorm epsilon precision issue, the other is a Transformers v5 meta-init incompatibility for old InternVL2 checkpoints." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Different code paths: non-persistent buffer corruption vs MoonshineStreaming sliding-window off-by-one." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "One is an FP8 accuracy regression; the other is an infinite-loop bug on a specific Qwen3_5 image input." + }, + { + "left": "issue:44327", + "right": "issue:47059", + "accept": false, + "reason": "QA pipeline top-k bounds crash and CLI completion attribute error are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Module import failure and tokenizer spacing behavior are not the same underlying bug." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer-related, but one is offset_mapping correctness and the other is add_special_tokens/BOS-EOS behavior for a specific model." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Offline cache loading and quantized model saving are different lifecycle paths with different failure modes." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Torch.compile attention-kernel divergence is unrelated to a docstring issue." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline tokenizer cache loading and wrong tokenizer decoder type in v5 are distinct regressions." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "The second is a broad multi-import regression, but it is not the same concrete import symbol breakage as the first." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "Both mention sam3, but one is a model-type warning and the other is a memory overflow; different bugs." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta and Sentence Transformers loading hangs are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Both involve tokenization, but one is BPE marker handling for a specific model and the other is a version-to-version tokenization change." + }, + { + "left": "issue:45381", + "right": "issue:46909", + "accept": false, + "reason": "Video position_ids correctness and a broken documentation example are not the same issue." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation kwargs handling after sync_gpus and D-FINE auxiliary-loss logic are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46909", + "accept": false, + "reason": "Qwen3OmniMoe video-input processing and a DIA documentation example are unrelated." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Custom model initialization breakage and NaNs after checkout/init are different failure modes." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Quantized model saving failure and pipeline-parallelism output corruption are separate bugs." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Both are runtime/loading problems, but they affect different models, platforms, and symptoms." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "A buffer read error and tokenizer_config class mismatch are different serialization/loading issues." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version compatibility and incorrect LLM output under pipeline parallelism do not share a code path." + }, + { + "left": "issue:43329", + "right": "issue:47059", + "accept": false, + "reason": "Multimodal token counting bug and CLI completion attribute error are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Saving dequantized Ministrals/Devstrals and loading Phi-4 multimodal are different model-loading/saving issues." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "A broken transformers env command and a sentence-transformers import NameError are unrelated." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE-layer handling and Qwen3-Next tied-weights warning are different model-specific bugs." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Wrong offset_mapping and inability to load cjvt/sleng-bert are not the same underlying tokenizer bug." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing export and a broad set of import errors after upgrade are related only at a very high level, not as one duplicate bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 51, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:29159", + "issue:36296", + "issue:39401", + "issue:42175", + "issue:42371", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43493", + "issue:43784", + "issue:43824", + "issue:43937", + "issue:44038", + "issue:44117", + "issue:44188", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44829", + "issue:44938", + "issue:44945", + "issue:45200", + "issue:45216", + "issue:45290", + "issue:45478", + "issue:45600", + "issue:45753", + "issue:46172", + "issue:46326", + "issue:46328", + "issue:46614", + "issue:46762" + ], + "soft_pairs": [ + "issue:42371|issue:44263", + "issue:44117|issue:44704", + "issue:42907|issue:43937", + "issue:39401|issue:44521", + "issue:43784|issue:45600", + "issue:46172|issue:46762", + "issue:44625|issue:44829", + "issue:44704|issue:45290", + "issue:44360|issue:44485", + "issue:43122|issue:44521", + "issue:43122|issue:46326", + "issue:43408|issue:43475", + "issue:44038|issue:44297", + "issue:42907|issue:43452", + "issue:22355|issue:29159", + "issue:44038|issue:44295", + "issue:45216|issue:45478", + "issue:44038|issue:44222", + "issue:39401|issue:46614", + "issue:43824|issue:44351", + "issue:43116|issue:45200", + "issue:43122|issue:44779", + "issue:42757|issue:44351", + "issue:36296|issue:43116", + "issue:43278|issue:44945", + "issue:22355|issue:29127", + "issue:42175|issue:44938", + "issue:43493|issue:45753", + "issue:44188|issue:46328" + ], + "prepared_review_unit_hash": "a240b198c28f7e3caff2fa2f2ee3c194d32ba723acd690cc6e400aeb1f74d283", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35534, + "estimated_input_tokens": 8884, + "estimated_eval_tokens": 18024 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the items span unrelated tokenizer, model-loading, training, packaging, multimodal, and parallelism bugs. None of the soft-similarity pairs look like the same underlying defect, so I would not merge any of them into a duplicate cluster.", + "confidence": 0.89, + "canonical_issue_reason": "No canonical issue: the cluster is heterogeneous and none of the items appear to describe the same underlying bug.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a strong global representative; the set spans unrelated subsystems and model families.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 backend settings and a GlmMoeDsaIndexer split bug are unrelated." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and AutoProcessor kwargs forwarding are different code paths." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Saving dequantized models and invalid GenerationConfig handling are distinct failures." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Qwen3 offset_mapping bugs and multimodal assistant mask generation are not the same issue." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "An import NameError in sentence-transformers and removed Sam3LiteText configs are unrelated." + }, + { + "left": "issue:46172", + "right": "issue:46762", + "accept": false, + "reason": "Deepseek-v4 dtype/transpose behavior and Minimax numerical mismatch are different model-specific problems." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "num_labels propagation and flash_attention_3 degenerate training do not share the same root cause." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "cached_file kwarg forwarding and apply_chat_template tool-call crashes affect different paths." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer version drift and multimodal assistant mask bugs are not the same defect." + }, + { + "left": "issue:43122", + "right": "issue:46326", + "accept": false, + "reason": "Both involve tokenizers, but one is a version regression and the other is a continue_final_message bug in mistral_common tokenizers." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "A model-type warning and a missing fpn_position_embeddings attribute are different SAM3 issues." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe loading and tokenizer_class mismatch on save_pretrained are separate problems." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Dequantized save failures and gguf_file from_pretrained breakage are not the same bug." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "Missing transformers.onnx and slow/fast tokenizer inconsistency are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe loading issues and position_ids buffer access are different code paths." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5-related, but one is save_pretrained corruption and the other is from_pretrained loading error; not enough to treat as one bug." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Qwen3-VL-Moe model loading and FP8 moe save_pretrained issues are distinct." + }, + { + "left": "issue:39401", + "right": "issue:46614", + "accept": false, + "reason": "Qwen3 offset_mapping and Wav2Vec2 phonemizer language-state bugs are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Missing Qwen2_5_VLForConditionalGeneration and missing HybridCache imports are different import regressions." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example-script multi-label output failure and Gemma 4 mm_token_type_ids defaults are different symptoms." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "A tokenizer version regression and a DeepSeek v5 tokenizer mismatch are not clearly the same root cause." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Hugging Face Hub import failure and transformers HybridCache import failure are separate dependency issues." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor parallel training bugs and a classification example script returning empty results are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "BF16-to-FP32 dtype drift and pipeline parallelism output corruption are different training/runtime issues." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "An import error for transformers.onnx and a LayoutLMv3 error-message clarity issue are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Missing tensorflow in the torch extra and Python 3.14 import failure are unrelated packaging/runtime issues." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "SigLIP2 implementation mismatch and Qwen3_5 infinite-loop behavior are not the same bug." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "torch.compile attention-kernel divergence and a slow padding inefficiency do not share a concrete code-path 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": 26, + "review_unit_index": 52, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28348", + "issue:33453", + "issue:38617", + "issue:39401", + "issue:42371", + "issue:42673", + "issue:42796", + "issue:42898", + "issue:43066", + "issue:43116", + "issue:43493", + "issue:43611", + "issue:43618", + "issue:43825", + "issue:43874", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44355", + "issue:44485", + "issue:44610", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:44995", + "issue:45072", + "issue:45083", + "issue:45092", + "issue:45278", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:46002", + "issue:46231", + "issue:46614", + "issue:46640", + "issue:46697", + "issue:46710", + "issue:46726", + "issue:46752", + "issue:46762", + "issue:46829", + "issue:46877", + "issue:46909", + "issue:46981" + ], + "soft_pairs": [ + "issue:46231|issue:46877", + "issue:42898|issue:46614", + "issue:43937|issue:44485", + "issue:38617|issue:45278", + "issue:43066|issue:44779", + "issue:44038|issue:45478", + "issue:43611|issue:46002", + "issue:43976|issue:45446", + "issue:42673|issue:45072", + "issue:44355|issue:46231", + "issue:45357|issue:45478", + "issue:44625|issue:46002", + "issue:43116|issue:43493", + "issue:44610|issue:46909", + "issue:45083|issue:45446", + "issue:42371|issue:43874", + "issue:46726|issue:46981", + "issue:44075|issue:44936", + "issue:43618|issue:46829", + "issue:43992|issue:44704", + "issue:43066|issue:46710", + "issue:43116|issue:43825", + "issue:28348|issue:46640", + "issue:46614|issue:46752", + "issue:44995|issue:46697", + "issue:42673|issue:42796", + "issue:45092|issue:46762", + "issue:33453|issue:39401" + ], + "prepared_review_unit_hash": "e64b89ad85fc540dbdbb0444a991304a935d4c29a3f48277303331d7e38534c7", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35735, + "estimated_input_tokens": 8934, + "estimated_eval_tokens": 18124 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are a loose collection of unrelated bug reports; none of the soft pairs look like true duplicates or the same fixable code-path issue. Most similarities are only at the model/tokenizer/Transformer-v5 theme level, not the concrete underlying bug.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:46614 is the strongest standalone candidate: it is open, clearly described, has inbound references, and reports a specific reproducible regression in Wav2Vec2PhonemeCTCTokenizer state handling.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46231", + "right": "issue:46877", + "accept": false, + "reason": "Docs parameter-name mismatch vs import-time docstring lint output; unrelated problems." + }, + { + "left": "issue:42898", + "right": "issue:46614", + "accept": false, + "reason": "v5 tokenization-space behavior change vs stale espeak backend language state; different tokenizer subsystems and bugs." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GenerationConfig validation error vs RoPE implementation discussion; not the same defect." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Single ImportError for one missing symbol vs many import errors after upgrade; too broad and not the same root cause." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Wrong decoder type in v5 vs DeepSeek tokenizer incorrect results; both tokenizer-related but different concrete failures." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3-VL-Moe bug report vs Qwen3.5 Moe from_pretrained error; same family, different code paths and symptoms." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "base_model_prefix loading regression vs NaNs in classification heads; different model-loading versus initialization issues." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version compatibility regression vs incorrect PyTorch version check in flex_attention; unrelated version-gating bugs." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in dataloader threads vs dtype mismatch in bfloat16 inference; distinct runtime problems." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Compiled Python file errors vs documentation parameter-name typo; not the same bug." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "save_pretrained visual encoder key regression vs Qwen3.5 from_pretrained failure; related model family but different serialization/loading defects." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "num_labels propagation bug vs NaNs in classification heads; different mechanisms and fixes." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Empty multi-label example output vs SigLIP2 HF/JAX discrepancy; separate application-level and model-implementation issues." + }, + { + "left": "issue:44610", + "right": "issue:46909", + "accept": false, + "reason": "Model input-size mismatch vs broken documentation example; one is code behavior, the other docs." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Helper output-length calculation bug vs flex_attention import version check; unrelated components." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API-setting guidance vs missing image-patch helper in GLM46V; not the same issue." + }, + { + "left": "issue:46726", + "right": "issue:46981", + "accept": false, + "reason": "Missing vision backend handling vs repeated 429 retries in AutoProcessor; both processor-related but different failure modes." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "SGD args ignored vs trainer.evaluate() failing after train(); distinct training bugs." + }, + { + "left": "issue:43618", + "right": "issue:46829", + "accept": false, + "reason": "CLIPOutput attentions missing vs VideoPrism output field mis-assignment; different model outputs." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Missing embed_tokens.weight in UMT5Encoder vs AutoProcessor not forwarding kwargs; unrelated loading/dispatch bugs." + }, + { + "left": "issue:43066", + "right": "issue:46710", + "accept": false, + "reason": "Wrong tokenizer decoder type vs DeepSeek accuracy regression after upgrade; not the same underlying problem." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example script returning empty results vs pipeline translation-task error message; different user-facing behaviors." + }, + { + "left": "issue:28348", + "right": "issue:46640", + "accept": false, + "reason": "GPT2 flash-attention support request vs Flash Attention 2 issue; same broad topic only, but no concrete same bug/change evidence." + }, + { + "left": "issue:46614", + "right": "issue:46752", + "accept": false, + "reason": "Stale phonemizer backend state vs apply_chat_template crashing on empty list; unrelated tokenizer edge cases." + }, + { + "left": "issue:44995", + "right": "issue:46697", + "accept": false, + "reason": "Second-forward stale indexer cache crash vs image processor over-padding; different model and preprocessing bugs." + }, + { + "left": "issue:42673", + "right": "issue:42796", + "accept": false, + "reason": "VRAM leak in multi-threaded inference vs tokenizer BPE-marker handling; not the same defect." + }, + { + "left": "issue:45092", + "right": "issue:46762", + "accept": false, + "reason": "Remote-code/meta-init incompatibility vs numerical differences in Minimax attention; separate model-loading versus math-implementation issues." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression vs Qwen3 offset_mapping bug; both tokenizer-related but different root causes and scopes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 53, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42617", + "issue:42673", + "issue:43064", + "issue:43122", + "issue:43408", + "issue:43441", + "issue:43450", + "issue:43673", + "issue:43742", + "issue:43824", + "issue:43827", + "issue:43937", + "issue:43986", + "issue:44117", + "issue:44263", + "issue:44291", + "issue:44380", + "issue:44410", + "issue:44466", + "issue:44779", + "issue:44829", + "issue:44977", + "issue:45071", + "issue:45081", + "issue:45216", + "issue:45310", + "issue:45341", + "issue:45405", + "issue:45446", + "issue:45561", + "issue:45563", + "issue:45588", + "issue:45632", + "issue:46002", + "issue:46077", + "issue:46298", + "issue:46328", + "issue:46512", + "issue:46655", + "issue:46752", + "issue:46762", + "issue:46981", + "issue:47000", + "issue:47059" + ], + "soft_pairs": [ + "issue:44263|issue:46762", + "issue:43827|issue:45446", + "issue:43064|issue:43673", + "issue:42222|issue:43441", + "issue:43824|issue:46981", + "issue:43673|issue:46328", + "issue:42222|issue:47000", + "issue:42673|issue:46655", + "issue:43937|issue:45563", + "issue:42222|issue:46752", + "issue:44117|issue:45588", + "issue:45405|issue:45446", + "issue:44291|issue:45071", + "issue:44380|issue:46077", + "issue:44466|issue:46002", + "issue:39401|issue:44779", + "issue:44410|issue:45310", + "issue:43937|issue:44977", + "issue:43450|issue:43986", + "issue:45071|issue:47059", + "issue:45341|issue:46328", + "issue:45071|issue:46512", + "issue:44829|issue:45216", + "issue:45341|issue:45561", + "issue:45632|issue:46981", + "issue:44117|issue:45081", + "issue:43408|issue:43742", + "issue:42617|issue:45341", + "issue:38617|issue:41628", + "issue:43122|issue:46298" + ], + "prepared_review_unit_hash": "0946d85c9a971751dffae81cefe86fc914005b624ed703ebdf0abbcabe16bda2", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36732, + "estimated_input_tokens": 9183, + "estimated_eval_tokens": 18622 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is a heterogeneous mix of unrelated Transformers issues (tokenization, model loading, attention backends, training, processors, testing utils, and CLI/environment regressions). No soft pair looks like the same underlying bug or change, so all proposed duplicate links are rejected.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43122 is the broadest, most representative regression report in the set: a version-to-version tokenization change with cross-version comparison and inbound references, making it the best single issue anchor among otherwise diverse reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43122 is the most generally useful issue for the cluster because it captures a concrete v5 regression in core tokenizer behavior rather than a model-specific or peripheral failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44263", + "right": "issue:46762", + "accept": false, + "reason": "Different bugs in different code paths: GlmMoeDsaIndexer split handling vs Minimax MSA attention numerical mismatch." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() is unrelated to the flex_attention AuxRequest version-check import bug." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "FSDP2/PEFT optimizer state corruption and GenerationMixin cache loss are separate training vs generation failures." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Vitpose model breakage and Ministral flash-attention failure affect different models and subsystems." + }, + { + "left": "issue:43824", + "right": "issue:46981", + "accept": false, + "reason": "Qwen2.5-VL import error is unrelated to AutoProcessor retry behavior on HTTP 429." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Cache handling in generation and numpy-to-list overhead in feature padding are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:47000", + "accept": false, + "reason": "Vitpose model loading and RTDetr loss infeasibility are different issues." + }, + { + "left": "issue:42673", + "right": "issue:46655", + "accept": false, + "reason": "VRAM leakage in Qwen3ForCausalLM threads is unrelated to SmolVLM video padding shape errors." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "Invalid GenerationConfig and stale num_return_sequences warning are distinct generation-config problems." + }, + { + "left": "issue:42222", + "right": "issue:46752", + "accept": false, + "reason": "Vitpose model breakage is unrelated to empty-list chat template crashes." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and flash_attention s_aux=None crashes are separate failures." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "An unreleased PEFT version bump is unrelated to the flex_attention import version check." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument and broken PretrainedConfig type checking are different v5 regressions." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "GPT2 attention scaling being ignored and Gemma4 bidirectional masking are different attention semantics bugs." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "lm_head serialization inconsistency is unrelated to NaNs in classification heads." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and likely different code paths." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Missing Qwen3Next projections and Qwen3.5 MoE from_pretrained errors are related only by model family, not the same bug." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Invalid GenerationConfig and Qwen3.5 flash-attention generation failure are not the same underlying issue." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Batched video processor shape errors and AutoProcessor dependency crashes are different video-processing problems." + }, + { + "left": "issue:45071", + "right": "issue:47059", + "accept": false, + "reason": "PretrainedConfig typing regressions and hf hub CLI completion errors are unrelated." + }, + { + "left": "issue:45341", + "right": "issue:46328", + "accept": false, + "reason": "A testing_utils bug and SequenceFeatureExtractor.pad inefficiency do not describe the same defect." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "Type-checking breakage in v5.4.0 and stale config.dtype after from_config are different config issues." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "FlashAttention-3 training collapse and Qwen3.5 checkpoint serialization are separate regressions." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Both mention testing utils, but one is a small utility bug and the other is an xdist race on captured_info.txt." + }, + { + "left": "issue:45632", + "right": "issue:46981", + "accept": false, + "reason": "Remote-code cache path collisions and repeated 429 retries are unrelated filesystem vs networking issues." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES None handling and the Mistral regex patch crash are distinct tokenizer-loading failures." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "SAM3 model-type warning and MobileLLM key error are different model-loading problems." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A 3d_parallel.py runtime failure is unrelated to the testing_utils bug." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they concern different missing names and likely different export regressions." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Tokenizer behavior regression and noisy logging are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 54, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:27957", + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43208", + "issue:43352", + "issue:43441", + "issue:43577", + "issue:43673", + "issue:43874", + "issue:43931", + "issue:43937", + "issue:44075", + "issue:44077", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44496", + "issue:44829", + "issue:44877", + "issue:44933", + "issue:45071", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45362", + "issue:45464", + "issue:45478", + "issue:45656", + "issue:46178", + "issue:46209", + "issue:46231", + "issue:46328", + "issue:46617", + "issue:46762", + "issue:46894", + "issue:46981" + ], + "soft_pairs": [ + "issue:45362|issue:46328", + "issue:42222|issue:46178", + "issue:39401|issue:46209", + "issue:44877|issue:46231", + "issue:42371|issue:44261", + "issue:44484|issue:45310", + "issue:43937|issue:44496", + "issue:44261|issue:46762", + "issue:42371|issue:46617", + "issue:44933|issue:45278", + "issue:38617|issue:44220", + "issue:44410|issue:45478", + "issue:44075|issue:45656", + "issue:43066|issue:43122", + "issue:43208|issue:46894", + "issue:44263|issue:44485", + "issue:43097|issue:44493", + "issue:44077|issue:46981", + "issue:43673|issue:45341", + "issue:44483|issue:45464", + "issue:43931|issue:44410", + "issue:41628|issue:45071", + "issue:27957|issue:43352", + "issue:44484|issue:45478", + "issue:43441|issue:44829", + "issue:43012|issue:43577", + "issue:43874|issue:44261" + ], + "prepared_review_unit_hash": "a6006c2a0cbc7e0b6d462a0b72278172a8a344b5bcc89fc12b82576661dee15a", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35359, + "estimated_input_tokens": 8840, + "estimated_eval_tokens": 17936 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous: the soft pairs share broad Transformers themes (tokenizers, imports, model loading, FlashAttention, generation, etc.) but do not look like the same concrete bug/change. I found no convincing duplicate pairs to merge.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: it describes multiple import regressions after a version upgrade, making it the closest thing to a cluster representative among otherwise mostly unrelated issue reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative issue because it summarizes a wide, user-visible regression class and is more likely than the others to subsume neighboring import-related complaints.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45362", + "right": "issue:46328", + "accept": false, + "reason": "Different subsystems and failures: Qwen3.5 chat inference crash vs SequenceFeatureExtractor padding performance." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "VitPose model breakage is unrelated to the tensorflow_text backend mapping error." + }, + { + "left": "issue:39401", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer offset mapping bug and FP8/B200 inference failure are unrelated code paths." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Granite speech config loading and a documentation parameter-name mismatch are not the same bug." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 API guidance and Minimax MLA layernorm epsilon behavior are unrelated." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Save_pretrained sharding limit question vs Qwen3.5 MoE from_pretrained failure; no shared bug." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Invalid GenerationConfig and unrecognized model/config loading are different failures." + }, + { + "left": "issue:44261", + "right": "issue:46762", + "accept": false, + "reason": "Both mention numerical issues, but one is a layernorm-epsilon problem and the other is an attention-vs-reference mismatch; too different to merge." + }, + { + "left": "issue:42371", + "right": "issue:46617", + "accept": false, + "reason": "TF32 settings and trust_remote_code cache FileNotFoundError are unrelated." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing import from image_utils is not the same concrete bug as a broad post-upgrade import-regression report." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "Missing layer_type_validation import and _torch_extract_fbank_features are unrelated." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen model families and different failures: missing projections vs from_pretrained error." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "SGD arguments ignored is a different optimizer bug from deepspeed causing double optimizer steps." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both involve tokenizer behavior in v5, but one is an internal decoder-type issue and the other is a broader tokenization difference; not clearly the same concrete bug." + }, + { + "left": "issue:43208", + "right": "issue:46894", + "accept": false, + "reason": "xLSTM training blockers and CohereASR double-shift loss bug are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "torch.split return handling in GLM Moe indexing is unrelated to the GLM-5 RoPE implementation discussion." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder and unexpected position-id keys are different regressions." + }, + { + "left": "issue:44077", + "right": "issue:46981", + "accept": false, + "reason": "PatchTSMixer post_init policy and AutoProcessor 429 retry behavior are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "GenerationMixin cache during chunked_prefill and a testing_utils bug are unrelated." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both concern chat/completions, but one is a general endpoint acceptance issue and the other is a Qwen3.5 streaming inference failure; not the same bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch and qwen3next missing projections are distinct model-loading/model-architecture issues." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor import failure and PreTrainedConfig type-checking breakage are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "Both mention Flash Attention 2, but they affect different models and likely different support gaps." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Save sharding default question vs Qwen3.5 MoE from_pretrained error are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Ministral FlashAttention failure and flash_attention_3 degenerate training are different concrete issues." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "A compile-time bf16 warning and Blip2 dtype not propagating to qformer are different bugs." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image-patch helper in GLM46V and Minimax layernorm epsilon 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": 26, + "review_unit_index": 55, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:38617", + "issue:42757", + "issue:42886", + "issue:42890", + "issue:43012", + "issue:43064", + "issue:43097", + "issue:43208", + "issue:43452", + "issue:43531", + "issue:43673", + "issue:43873", + "issue:43874", + "issue:43931", + "issue:43986", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44987", + "issue:45003", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45440", + "issue:45588", + "issue:45753", + "issue:46050", + "issue:46123", + "issue:46172", + "issue:46298", + "issue:46369", + "issue:46655", + "issue:46664", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46821", + "issue:46897", + "issue:46981" + ], + "soft_pairs": [ + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:43931|issue:44038", + "issue:43097|issue:43452", + "issue:43986|issue:45341", + "issue:45003|issue:45341", + "issue:42886|issue:46981", + "issue:44261|issue:44485", + "issue:45753|issue:46369", + "issue:44117|issue:44987", + "issue:43531|issue:45341", + "issue:43097|issue:45588", + "issue:43673|issue:46123", + "issue:43874|issue:44263", + "issue:45440|issue:46172", + "issue:44410|issue:45375", + "issue:43986|issue:45372", + "issue:44038|issue:44410", + "issue:43097|issue:46050", + "issue:38617|issue:42757", + "issue:43064|issue:44315", + "issue:43208|issue:46897", + "issue:43986|issue:44484", + "issue:46655|issue:46664", + "issue:46298|issue:46752", + "issue:42890|issue:46298", + "issue:38617|issue:44351", + "issue:44062|issue:44117", + "issue:43097|issue:43873", + "issue:43012|issue:46821", + "issue:44117|issue:46701", + "issue:46172|issue:46710" + ], + "prepared_review_unit_hash": "de9ec07b50adcb2d60717b73d0d40bf15ece47cd3a2e543f7e9e47e4cb730e8f", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36513, + "estimated_input_tokens": 9129, + "estimated_eval_tokens": 18514 + }, + "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. None of the soft-pair candidates look like the same underlying bug or change, so all soft edges are rejected.", + "confidence": 0.76, + "canonical_issue_reason": "Issue #43873 is the most developed and actively discussed report in the set, so it is the best single anchor if one issue must represent the cluster; however, the cluster itself is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43873 has the richest context, is still open, and describes a concrete actionable regression, making it the strongest overall issue in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different bugs in DSA/xLSTM-style indexing vs DeepSeek-V4 tensor handling; no shared code path." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "Both mention DSA/indexing, but one is a missing ReLU in the indexer and the other is a dtype/transpose bug." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Same broader Qwen3-VL area, but one is a model weight-shape mismatch and the other is a Qwen3-VL-Moe bug." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "One is a removed config API in v5, the other is gguf loading failing in from_pretrained; different failure modes." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "AutoProcessor/torchvision import crash vs a testing_utils bug; unrelated code paths." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "sys.modules access in modeling_utils vs a testing helper bug are separate issues." + }, + { + "left": "issue:42886", + "right": "issue:46981", + "accept": false, + "reason": "Both involve Hub/network behavior, but offline cache loading and repeated 429 retries are distinct problems." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "MLA rms_norm_eps/config precision issue vs GLM-5 RoPE implementation; not the same root cause." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "Qwen3_5 infinite-loop bug on a specific image vs SAM3 memory overflow; different models and symptoms." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping None handling vs a specific transformers>=5.1 loading failure; separate loading bugs." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Sliding_window issue in Qwen3-MoE vs a testing utility bug; not the same bug." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "v5 API removal vs flash_attention crashing on sink-less models; distinct code paths." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "GenerationMixin cache missing during chunked_prefill vs partial-batch dropping in MaskGenerationPipeline; different defects." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Missing get_number_of_image_patches vs torch.split return values; separate attribute and tensor-shape bugs." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Native DeepseekV3MoE divergence vs a DeepSeek-V4 tensor/transposition issue; different implementations." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Missing attn projections in qwen3next vs a missing config field silently dropped; different layers and failure modes." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Video-model loading crash without torchvision vs Gemma 4 processor import error from mistral_common; unrelated dependencies." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen3-family, but one is a Qwen3-VL-Moe issue and the other is missing projections in qwen3next." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Removed API in v5 vs quantized model save/load corruption; not the same underlying bug." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different modules and symbols." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state bug vs Liger Kernel not applied with model_init; separate training-path issues." + }, + { + "left": "issue:43208", + "right": "issue:46897", + "accept": false, + "reason": "xLSTM training blockers vs Florence2 double-shift loss bug; different models and mechanics." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "AutoProcessor import crash vs a save_pretrained shard-size question; unrelated." + }, + { + "left": "issue:46655", + "right": "issue:46664", + "accept": false, + "reason": "Both are SmolVLM video-processor issues, but padding math and num_frames semantics are different bugs." + }, + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "Noisy logging message vs apply_chat_template crashing on an empty list; unrelated." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "SAM HQ test flakiness vs logging noise; not the same bug." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different public APIs and packages." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken special-kwarg conflict vs tokenizer-mapping None handling; separate tokenizer initialization bugs." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "v5 API removal vs quantization/offloading behavior; distinct code paths and symptoms." + }, + { + "left": "issue:43012", + "right": "issue:46821", + "accept": false, + "reason": "PyTorch warning during bfloat16 compilation vs TimesFM window_size AttributeError; unrelated." + }, + { + "left": "issue:44117", + "right": "issue:46701", + "accept": false, + "reason": "Tokenizer mapping fallback bug vs image-text-to-text pipeline failure for DiffusionGemmaGenerationOutput." + }, + { + "left": "issue:46172", + "right": "issue:46710", + "accept": false, + "reason": "DeepSeek-V4 tensor-construction bug vs an accuracy regression after upgrade; not the same root 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": 26, + "review_unit_index": 56, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:36683", + "issue:38617", + "issue:42175", + "issue:42490", + "issue:42757", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43502", + "issue:43550", + "issue:43582", + "issue:43824", + "issue:43825", + "issue:43906", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44016", + "issue:44077", + "issue:44162", + "issue:44222", + "issue:44368", + "issue:44380", + "issue:44483", + "issue:44484", + "issue:44514", + "issue:44655", + "issue:44704", + "issue:44936", + "issue:44938", + "issue:45103", + "issue:45478", + "issue:45479", + "issue:45584", + "issue:45712", + "issue:45753", + "issue:46132", + "issue:46172", + "issue:46178", + "issue:46298", + "issue:46365", + "issue:46421", + "issue:47059" + ], + "soft_pairs": [ + "issue:43502|issue:43992", + "issue:44704|issue:45103", + "issue:44936|issue:46365", + "issue:44380|issue:46421", + "issue:44938|issue:46178", + "issue:43986|issue:44016", + "issue:42907|issue:45753", + "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:43825|issue:44655", + "issue:43352|issue:44380", + "issue:43352|issue:43976", + "issue:43066|issue:43906", + "issue:43097|issue:46298", + "issue:45584|issue:46365", + "issue:44380|issue:46132", + "issue:44162|issue:46172", + "issue:42890|issue:47059", + "issue:44077|issue:46365", + "issue:44222|issue:45478", + "issue:36683|issue:43582", + "issue:44380|issue:45479", + "issue:42757|issue:43824", + "issue:43064|issue:44380", + "issue:42175|issue:46365", + "issue:42490|issue:43012", + "issue:44483|issue:44514", + "issue:38617|issue:43502" + ], + "prepared_review_unit_hash": "ac49c987367b0b10b22237e49e9d54c85758325c435c94e2119fe3114bf33f4b", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37226, + "estimated_input_tokens": 9307, + "estimated_eval_tokens": 18870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The soft pairs are mostly superficial similarity matches across unrelated bugs (imports, warnings, saving, backend behavior, and model-specific failures). None of the candidate pairs look like the same underlying issue or a mergeable PR duplicate.", + "confidence": 0.83, + "canonical_issue_reason": "issue:44380 is the most connected node in the soft-pair graph and is a well-scoped concrete runtime bug, making it the best representative issue among a very heterogeneous set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44380 has the highest apparent centrality and describes a specific code-path regression, so it is the most suitable global representative even though the cluster is not truly duplicate-cohesive.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "One is a local-files-only network/request bug; the other is a missing model weight during from_pretrained. Different failure modes and code paths." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and auto-docstring parsing with future annotations are unrelated subsystems." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "Trainer.evaluate() after train() is not the same as an import failure caused by datasets shadowing Trainer imports." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "Attention scaling being ignored under SDPA/FlashAttention is unrelated to cache lazy initialization and chunked prefill recompilation." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python 3.14 import compatibility and an invalid backend mapping for tensorflow_text are different root causes." + }, + { + "left": "issue:43986", + "right": "issue:44016", + "accept": false, + "reason": "Missing torchvision import handling is unrelated to a notebook syntax error." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Saving dequantized models and an image-triggered infinite generation loop are different bugs." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline save_pretrained failure and dequantized model save failure are both save-related but not the same underlying bug." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A warning about bfloat16/float32 precision is not the same as a torch.compile SDPA failure on a specific model." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Dead dummy classes in a repo audit are unrelated to noisy logging output." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Both mention warnings around embeddings, but one is a PyTorch precision warning and the other is a model/config warning about tie_word_embeddings." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing embed_tokens.weight during load is unrelated to the default max_shard_size in save_pretrained." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training blockers are model-specific and do not match a Bamba compile/SDPA regression." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "A misleading pipeline error message and inability to save pipeline objects are different issues." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "A model explicitly not supporting Flash Attention 2.0 is not the same as GPT2 ignoring attention scaling under backends." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "A model/backend capability issue is unrelated to Python version compatibility during import." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type in v5 and an isolated reproduction of another issue do not describe the same bug." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "A removed API with no indication is unrelated to a noisy logging message." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation failure and a Trainer import conflict are unrelated." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "Attention scaling behavior under SDPA/FlashAttention is not the same as register() mutating a function's behavior." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage and a DeepSeek tensor transpose/dtype bug are model-specific and distinct." + }, + { + "left": "issue:42890", + "right": "issue:47059", + "accept": false, + "reason": "Flaky test seeding and a cli incompatibility with huggingface_hub 1.22.0 are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "patchtsmixer post_init validation and an import conflict with datasets have no shared underlying code path." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "FP8 save_pretrained behavior and a Qwen3.5 MoE from_pretrained error are different model lifecycle bugs." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config missing vocab_size and an Apple Silicon TypeError in caching_allocator_warmup are unrelated." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention scaling being ignored is unrelated to a degenerate zero-loss classification setup." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "ImportError for is_offline_mode and missing Qwen2_5_VLForConditionalGeneration are different missing-symbol problems." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT2 attention scaling under SDPA/FlashAttention." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "Backend packaging for tensorflow versus Trainer import failure due to datasets are both dependency issues but not the same bug." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "A trailing-slash local model warning and a bfloat16 compile precision warning are different concerns." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Chat/completions request rejection and batched chat template crashes are different endpoints and failure modes." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "A missing configuration_utils import and unexpected API requests despite local_files_only=True are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 57, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:29870", + "issue:33453", + "issue:36010", + "issue:36296", + "issue:36683", + "issue:42757", + "issue:42994", + "issue:43010", + "issue:43064", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43525", + "issue:43526", + "issue:43611", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43867", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44077", + "issue:44230", + "issue:44355", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44492", + "issue:44625", + "issue:44655", + "issue:44908", + "issue:44938", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45290", + "issue:45310", + "issue:46050", + "issue:46077", + "issue:46082", + "issue:46365", + "issue:46851", + "issue:46877", + "issue:47059" + ], + "soft_pairs": [ + "issue:43618|issue:44380", + "issue:43756|issue:44485", + "issue:43208|issue:43526", + "issue:43618|issue:43761", + "issue:45276|issue:46077", + "issue:36010|issue:42757", + "issue:43299|issue:43867", + "issue:42994|issue:44655", + "issue:44038|issue:44230", + "issue:44483|issue:45290", + "issue:44625|issue:46365", + "issue:43352|issue:43937", + "issue:44355|issue:46877", + "issue:36683|issue:43475", + "issue:43611|issue:44655", + "issue:44625|issue:46077", + "issue:44938|issue:47059", + "issue:44655|issue:46050", + "issue:43976|issue:44938", + "issue:43867|issue:45216", + "issue:36683|issue:44077", + "issue:46077|issue:46082", + "issue:43618|issue:45216", + "issue:36683|issue:43525", + "issue:36683|issue:44625", + "issue:29870|issue:36296", + "issue:43064|issue:44492", + "issue:43010|issue:45278", + "issue:44075|issue:46365", + "issue:43992|issue:45310", + "issue:44908|issue:46851", + "issue:29870|issue:33453" + ], + "prepared_review_unit_hash": "dfb30089cd59dd2952e9369303327e181d5715c0d1501fed0ab665c6294d1131", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36877, + "estimated_input_tokens": 9220, + "estimated_eval_tokens": 18696 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated regressions, import/version incompatibilities, and model-specific bugs. I don\u2019t see a true duplicate cluster here; the soft similarities are mostly broad subsystem overlap rather than the same underlying issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:36683 is the most active concrete bug report here and a reasonable anchor for the set, with strong discussion activity and a clearly described model/config regression.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:36683 is the best single issue to represent the cluster because it is open, highly discussed, and more concrete than the surrounding environment/import-only reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "Both involve attention behavior, but one is missing CLIP attentions while the other ignores GPT-2 attention scaling config; different code paths and fixes." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Both mention RoPE, but they describe different model implementations and different symptoms; not the same underlying bug." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "Completely unrelated: xLSTM training bugs versus a BeitImageProcessorFast label-reduction issue." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Both are CLIP v5 regressions, but one loses attentions and the other returns hidden_states=None; related theme, not clearly the same bug." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Different Gemma4 problems: token embedding resize propagation versus bidirectional-attention mask construction." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they come from different modules and different missing symbols; separate compatibility issues." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Both are model-loading failures, but for different failure modes and likely different root causes." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Both involve saving, but quantized-model save failure and Pipeline.save_pretrained failure are different problems." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Same family of Qwen3-VL-Moe topics, but one is a transformers regression and the other is fp8 inference support; not the same change." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "Different APIs and symptoms: chat/completions request handling versus apply_chat_template crashing on tool-call assistant messages." + }, + { + "left": "issue:44625", + "right": "issue:46365", + "accept": false, + "reason": "Unrelated: Qwen3.5 num_labels propagation versus a Trainer import conflict with datasets." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Different models and failures: Flash Attention 2 support gating versus an invalid GenerationConfig issue." + }, + { + "left": "issue:44355", + "right": "issue:46877", + "accept": false, + "reason": "One is about compiled Python files failing, the other about import-time docstring lint output; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Different missing attributes on different model outputs; same broad pattern, but not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Base-model-prefix loading regression and Pipeline.save_pretrained failure are separate issues." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Different Gemma4/Qwen3.5 bugs with no shared code-path indicated." + }, + { + "left": "issue:44938", + "right": "issue:47059", + "accept": false, + "reason": "Python 3.14 import failure and a huggingface_hub CLI compatibility failure are different environment breakages." + }, + { + "left": "issue:44655", + "right": "issue:46050", + "accept": false, + "reason": "Both are save/load related, but one is Pipeline serialization and the other is sinq quantization round-trip failure." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Both are version-compatibility problems, but they target different Python releases and likely different causes." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "One is about sorted state_dict loading order, the other about Qwen3.5 save_pretrained checkpoint correctness." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Unrelated: missing Gemma3Config attribute versus a patchtsmixer post_init policy change." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Different model configuration validations: attention-mask behavior versus head_dim constraints." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIP output regression and Qwen3.5 save_pretrained regression are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Different missing config attributes on different models; not the same bug." + }, + { + "left": "issue:36683", + "right": "issue:44625", + "accept": false, + "reason": "Gemma3Config vocab_size and Qwen3.5 num_labels are unrelated configuration issues." + }, + { + "left": "issue:29870", + "right": "issue:36296", + "accept": false, + "reason": "Bart generate() UnboundLocalError and tensor-parallel training bug are different failures." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "Training optimizer-state corruption and a typo in cache strategies are unrelated." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "A no_grad decoration issue in cache updates is not the same as broad import failures after upgrading." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "SGD argument handling and Trainer import conflicts are unrelated." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Both are from_pretrained/load issues, but they involve different models and distinct missing-weight behaviors." + }, + { + "left": "issue:44908", + "right": "issue:46851", + "accept": false, + "reason": "Scheduler kwargs are ignored versus variable-length Qwen3.5 position_ids handling; unrelated." + }, + { + "left": "issue:29870", + "right": "issue:33453", + "accept": false, + "reason": "Bart generate() decoder_attention_mask error and tokenizer-loading regression are separate bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 58, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28282", + "issue:30333", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42890", + "issue:42898", + "issue:43010", + "issue:43064", + "issue:43065", + "issue:43502", + "issue:43575", + "issue:43576", + "issue:43618", + "issue:43723", + "issue:43824", + "issue:43867", + "issue:43874", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44077", + "issue:44112", + "issue:44222", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44387", + "issue:44393", + "issue:44485", + "issue:44556", + "issue:44734", + "issue:44855", + "issue:44964", + "issue:45125", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45850", + "issue:46132", + "issue:46202", + "issue:46285", + "issue:46617", + "issue:46664", + "issue:47000" + ], + "soft_pairs": [ + "issue:44393|issue:47000", + "issue:43867|issue:46202", + "issue:44355|issue:44855", + "issue:36683|issue:43064", + "issue:43867|issue:45375", + "issue:42890|issue:45561", + "issue:43010|issue:45850", + "issue:36296|issue:43575", + "issue:41669|issue:44246", + "issue:43618|issue:44077", + "issue:43010|issue:43976", + "issue:43867|issue:45357", + "issue:43992|issue:45478", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:44360|issue:45440", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:43502|issue:44062", + "issue:42898|issue:46664", + "issue:44734|issue:46617", + "issue:43723|issue:44556", + "issue:44222|issue:45125", + "issue:43502|issue:44485", + "issue:43824|issue:46617", + "issue:43010|issue:43065", + "issue:43874|issue:46617", + "issue:43576|issue:46132", + "issue:44387|issue:44556", + "issue:30333|issue:33453", + "issue:45125|issue:45478" + ], + "prepared_review_unit_hash": "43f2d506d8f4dd62322e69b9ecb785a3433715c43a989441ec0feeaaa721a3e9", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36694, + "estimated_input_tokens": 9174, + "estimated_eval_tokens": 18604 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Heterogeneous cluster with a loose concentration around v5 loading/regression complaints, but no soft pair is mergeable; the previously overcalled 44393/47000 pair is not accepted.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43723 is the broadest user-facing loading-regression representative in the set and best captures the cluster\u2019s most common theme without overfitting to a niche model family.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43723 is the strongest global issue choice because it is generic, high-signal, and closest to the dominant loading/failure pattern across the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44393", + "right": "issue:47000", + "accept": false, + "reason": "Different model families and failure modes: Qwen3-VL 2D box hallucination vs RTDetrLoss cost-matrix infeasibility crash." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "Both are load/roundtrip-related at a high level, but the concrete bugs differ: sorted state_dict loading vs Gemma pipeline roundtrip on a token character." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Unrelated import/runtime failures: compiled Python file handling vs a Python 3.13 TorchScript parse/IndentationError." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Different subsystems and symptoms: missing Gemma3Config attribute vs distributed optimizer-state corruption under FSDP2/PEFT." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "Both touch loading, but one is a state_dict sorting issue and the other is a missing vision-config field being dropped." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Both are test-flake reports, but one is missing set_seed() in integration tests and the other is an xdist file race." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "No_grad decoration request and broad v5 breakage complaint are different problem scopes and code paths." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel is the only overlap; one is a generic training bug, the other is a model-specific OOM when loading Qwen2-57B-A14B-Instruct." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Both mention import slowness, but there is no evidence they share the same root cause or touched code path." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "Different fixes: CLIPOutput attentions assignment vs patchtsmixer post_init policy." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Completely different concerns: cache/layer no_grad decoration vs Python-version compatibility." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Generic load/save regression vs Qwen3.5 visual-encoder key regression; not the same concrete bug." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Both are from_pretrained failures, but they involve different Qwen3.5 models and different loading-path defects." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch dependency import error is unrelated to tokenizer loading regression." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "Both are CLIP-related, but the concrete structural changes differ and do not point to one shared bug." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer/ReLU discussion is unrelated to DeepSeekV3 implementation divergence." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Test-seed flakiness and multimodal model loading failure are different issues." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Both are CI/test-related, but one is about missing set_seed() and the other about a stale device override test." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Offline-loading network leakage and AddedToken keyword duplication are unrelated bugs." + }, + { + "left": "issue:42898", + "right": "issue:46664", + "accept": false, + "reason": "Tokenization cleanup semantics and video num_frames semantics are different areas and different failure modes." + }, + { + "left": "issue:44734", + "right": "issue:46617", + "accept": false, + "reason": "KV-cache continuation tensor-indexing crash and trust_remote_code cache FileNotFoundError are different code paths." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Both are loading regressions, but tokenizer loading in v5 and checkpoint reload compatibility after version upgrade are not the same bug." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained behavior and missing tensor-parallel plan are distinct model concerns." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Local-files-only network requests and GLM-5 RoPE implementation discussion are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:46617", + "accept": false, + "reason": "Missing exported class import and cache FileNotFoundError are different failure modes." + }, + { + "left": "issue:43874", + "right": "issue:46617", + "accept": false, + "reason": "Missing image-processor method and cache FileNotFoundError are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Broken transformers env command and AttentionInterface.register behavior changes are different issues." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 quantization memory growth/OOM and reload-after-upgrade failure are separate problems." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow reporting on failed training is unrelated to tokenizer loading regression." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is missing _tp_plan and the other is a from_pretrained error; not the same underlying bug." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the analyst overstates the only accepted soft pair. In the packet, issue:44393 (Qwen3-VL 2D bounding box hallucination/error) and issue:47000 (RTDetrLoss cost-matrix infeasible crash) are different model families and failure modes, and the packet itself marks the pair as deterministic_accept=false. The summary also says \u2018only one soft pair looks like the same underlying bug,\u2019 which is not grounded by the packet since none of the soft pairs are supported as mergeable." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 59, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "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:43232", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43576", + "issue:43673", + "issue:43824", + "issue:43827", + "issue:43901", + "issue:44222", + "issue:44230", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44749", + "issue:44961", + "issue:45071", + "issue:45125", + "issue:45362", + "issue:45375", + "issue:45468", + "issue:45542", + "issue:45850", + "issue:46143", + "issue:46144", + "issue:46209", + "issue:46617", + "issue:46664", + "issue:46726", + "issue:47059" + ], + "soft_pairs": [ + "issue:42548|issue:44351", + "issue:44355|issue:44749", + "issue:30990|issue:31795", + "issue:45125|issue:45375", + "issue:43502|issue:44222", + "issue:43502|issue:44961", + "issue:43827|issue:44961", + "issue:46664|issue:46726", + "issue:43010|issue:43232", + "issue:30333|issue:36296", + "issue:29870|issue:31795", + "issue:43299|issue:44556", + "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:44315|issue:44360", + "issue:43576|issue:43824", + "issue:44961|issue:45468", + "issue:44485|issue:44961", + "issue:44351|issue:46617", + "issue:36296|issue:43901", + "issue:43352|issue:44961", + "issue:43475|issue:46617", + "issue:46143|issue:46144", + "issue:44222|issue:44230", + "issue:41628|issue:42548", + "issue:44315|issue:46209", + "issue:45362|issue:45850", + "issue:43673|issue:44492", + "issue:44485|issue:46664", + "issue:44273|issue:45542" + ], + "prepared_review_unit_hash": "033262e1e006fb94563d2a2e13fa292b045d92d494ce67409285d070e690586c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37262, + "estimated_input_tokens": 9316, + "estimated_eval_tokens": 18888 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster looks heterogeneous: most soft-similarity pairs are cross-topic false positives spanning unrelated import errors, model-loading regressions, docs issues, and model-specific bugs. I would not merge any of the proposed soft edges.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45850 is the broadest umbrella ('Since 5.0 version it breaks too many models'), but it is still too vague to serve as a strong canonical duplicate target for this mixed set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols and likely different breakages (`PreTrainedModel` vs `HybridCache`)." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "One is about compiled Python files, the other about a training slowdown after upgrade; different bugs." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Stuck loading is a runtime problem; the other is documentation confusion about `forward` arguments." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Tensor-parallel `_tp_plan` support and a missing config field are different model integration issues." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "`local_files_only` network leakage and FP8 `save_pretrained` are unrelated behaviors." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No shared underlying bug; `racoon` is unrelated to offline loading." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing `pipeline()` vs an unrelated `racoon` issue." + }, + { + "left": "issue:46664", + "right": "issue:46726", + "accept": false, + "reason": "Different video/image processor semantics; not the same bug." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "`no_grad` decoration on cache updates and generation kwargs handling are distinct generation concerns." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow job status reporting and tensor-parallel training are unrelated." + }, + { + "left": "issue:29870", + "right": "issue:31795", + "accept": false, + "reason": "Generation-time `decoder_attention_mask` crash is not the same as forward-doc confusion." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Both mention v5/v5.x regressions, but one is Qwen3-VL loading and the other is checkpoint reload compatibility; not the same concrete bug." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Both are tokenizer/model loading regressions, but for different model families and failure modes." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file execution errors and `PretrainedConfig` type-checking breakage are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Both are import/CLI-related v5 issues, but one is Hub import resolution and the other is the `transformers env` command." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "CLI breakage and config type-checking are different areas and not mergeable as one fix." + }, + { + "left": "issue:43576", + "right": "issue:47059", + "accept": false, + "reason": "`transformers env` CLI regression and unrelated huggingface_hub completion breakage are different code paths." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion in model.forward is not the same as a missing `layer_type_validation` import." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel not applied in `model_init` and a DSA indexer discussion are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI breakage and missing `Qwen2_5_VLForConditionalGeneration` import are different regressions." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "`racoon` is unrelated to Gemma-4 audio positional encoding." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "GLM-5 RoPE implementation and `racoon` are unrelated." + }, + { + "left": "issue:44351", + "right": "issue:46617", + "accept": false, + "reason": "Both are cache/loading-related import problems, but they concern different missing behaviors and likely different modules." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor-parallel training and text-classification docs behavior are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:44961", + "accept": false, + "reason": "Flash Attention support for Nemotron and `racoon` share no underlying bug." + }, + { + "left": "issue:43475", + "right": "issue:46617", + "accept": false, + "reason": "SAM 3 Video attribute error and trust-remote-code cache loading are different failures." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Missing `**kwargs` pass-through and attention-interface usage are related RoFormer topics but not the same concrete bug." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "FP8 save/load bug and FP8 inference support request are different problems." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Missing `AutoImageProcessor` and missing `PreTrainedModel` imports are distinct export regressions." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger Kernel application in `model_init` and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Qwen3.5-35B chat crash is model-specific; the other is a broad v5 breakage complaint." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Generation cache missing during chunked prefill and a cache-strategy typo are not the same bug." + }, + { + "left": "issue:44485", + "right": "issue:46664", + "accept": false, + "reason": "GLM-5 RoPE implementation and SmolVLM frame-count semantics are unrelated." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading failure and undefined TensorFlow backend error 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": 26, + "review_unit_index": 60, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:31795", + "issue:33453", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46218", + "issue:46328", + "issue:46655", + "issue:46664", + "issue:47059" + ], + "soft_pairs": [ + "issue:44623|issue:45230", + "issue:44038|issue:45125", + "issue:44273|issue:45003", + "issue:44360|issue:46664", + "issue:41669|issue:43673", + "issue:44393|issue:45071", + "issue:44393|issue:45753", + "issue:43576|issue:44556", + "issue:43525|issue:45850", + "issue:44556|issue:45092", + "issue:43572|issue:45850", + "issue:44492|issue:45850", + "issue:44230|issue:46095", + "issue:45850|issue:47059", + "issue:45278|issue:45850", + "issue:42371|issue:44492", + "issue:41669|issue:46218", + "issue:43408|issue:44315", + "issue:44246|issue:44273", + "issue:42890|issue:44908", + "issue:43576|issue:44908", + "issue:43352|issue:44393", + "issue:43976|issue:44393", + "issue:44908|issue:45071", + "issue:44295|issue:45230", + "issue:44297|issue:45230", + "issue:44492|issue:46328", + "issue:41669|issue:44351", + "issue:44704|issue:44908", + "issue:41669|issue:45341", + "issue:43976|issue:46218", + "issue:44077|issue:44861", + "issue:31795|issue:33453", + "issue:43576|issue:44273", + "issue:44315|issue:44829", + "issue:43519|issue:46655", + "issue:44360|issue:45468", + "issue:44273|issue:44908", + "issue:45230|issue:45310" + ], + "prepared_review_unit_hash": "0eed20966a3a48c028dfb7d9e2d220e528936e20300fd57c387c3d7951c577a3", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 38824, + "estimated_input_tokens": 9706, + "estimated_eval_tokens": 19668 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab-bag of unrelated Transformers regressions and model-specific bugs. Only one soft duplicate looks plausible: the lazy-loading/import path issues around issue 44273 and issue 45003.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44273 is the broadest and most central report in the only coherent thread here (lazy loading/import behavior), and it has multiple nearby soft-similarity candidates.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44273 is the best issue anchor: open, discussion-heavy, and directly names the shared failure mode the closest duplicate candidate appears to target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "'processor.save_pretrained(...) missing files' and a generic 'Bug report' are too underspecified to be the same bug." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Both involve Qwen3.5/Moe, but one is a general bug report and the other is a missing _tp_plan tensor-parallelism issue; different failure modes." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": true, + "reason": "Both point to lazy loading/import machinery failing; unsafe sys.modules access in modeling_utils is a plausible root cause for 'lazy loading is not working properly'." + }, + { + "left": "issue:44360", + "right": "issue:46664", + "accept": false, + "reason": "DSA indexer/ReLU discussion is unrelated to SmolVLM video processor frame-count semantics." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-* cleanup/performance and GenerationMixin cache chunked-prefill are different code paths and symptoms." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL bbox output problems are unrelated to PretrainedConfig type checking regressions." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Same model family, but 2D bbox hallucinations and infinite loop on a specific image are distinct concrete bugs." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken 'transformers env' CLI and checkpoint reload failures after v4.57->v5.x are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "Missing pad_token_id on Llama4Config is a specific config bug, not the broad 'v5 breaks too many models' report." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are v5 upgrade regressions, but one is general checkpoint reload and the other is InternVL2 remote-code/meta-init incompatibility." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "StableLm pad_token_idx regression is model-specific and not the same as the broad version-5 breakage complaint." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "A cache-strategy typo is unrelated to generic v5 model breakage." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5 Moe is not the same as a weight conversion-script question." + }, + { + "left": "issue:45850", + "right": "issue:47059", + "accept": false, + "reason": "Broad v5 model breakage and a cli test failure with huggingface_hub 1.22.0 are unrelated." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "Both mention many breakages after upgrade, but they are different umbrella reports rather than the same underlying bug." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API guidance and a cache-strategies typo are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import-time slowdown from import-* usage and a specific Qwen3_5 module import failure are different problems." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "sam3_video/sam3_tracker warning and Liger Kernel not applied with model_init are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "'import transformers takes long sometimes' is a performance symptom, but not clearly the same concrete bug as lazy loading not working properly." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Missing set_seed in tests and inverse_sqrt scheduler kwargs handling are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken env CLI and scheduler kwargs behavior are different subsystems." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention 2 unsupported on NemotronH and Qwen3-VL bbox hallucinations are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility and Qwen3-VL output errors are unrelated." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug and PretrainedConfig type checking regression are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "position_ids buffer access error and a generic bug report are not the same issue." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "tokenizer.save_pretrained class-name mismatch is unrelated to a generic bug report." + }, + { + "left": "issue:44492", + "right": "issue:46328", + "accept": false, + "reason": "Cache strategy typo and SequenceFeatureExtractor.pad inefficiency are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "import-* cleanup/perf and missing HybridCache import are different import-path bugs." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs are different APIs and code paths." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Model import performance and a testing_utils bug are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python compatibility and Qwen3_5 import failure are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "patchtsmixer post_init policy and _get_tied_weight_keys crashing on list-typed keys are not the same bug." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion in model.forward and tokenizer loading regression are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken CLI env command and lazy loading not working are different failures." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger Kernel application in model_init and flash_attention_3 degenerate training are different bugs." + }, + { + "left": "issue:43519", + "right": "issue:46655", + "accept": false, + "reason": "Incorrect timestamp calculation and incorrect video padding calculation are different processor bugs." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer/ReLU discussion is unrelated to Gemma4AudioRelPositionalEncoding." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report and Qwen3.5 Moe from_pretrained error are not the same concrete issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft edge (44273 \u2194 45003) is framed as plausible rather than certain, and the other rejections are appropriately cautious and specific to distinct failure modes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 61, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:28348", + "issue:29870", + "issue:30333", + "issue:31515", + "issue:31795", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:42936", + "issue:43066", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43572", + "issue:43576", + "issue:43618", + "issue:43668", + "issue:43704", + "issue:43717", + "issue:44031", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44351", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44861", + "issue:44908", + "issue:44938", + "issue:44998", + "issue:45250", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46144", + "issue:46172", + "issue:46218", + "issue:46894", + "issue:47059" + ], + "soft_pairs": [ + "issue:43519|issue:44485", + "issue:45468|issue:46894", + "issue:43576|issue:46218", + "issue:44655|issue:44861", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:44485|issue:45468", + "issue:43704|issue:44485", + "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:44351|issue:46218", + "issue:43352|issue:46218", + "issue:42913|issue:43066", + "issue:44861|issue:47059", + "issue:28348|issue:45250", + "issue:44908|issue:44998", + "issue:44492|issue:44998", + "issue:44998|issue:45850", + "issue:43618|issue:44998", + "issue:44938|issue:44998", + "issue:44861|issue:45290", + "issue:31515|issue:38617", + "issue:43572|issue:44861", + "issue:44230|issue:44623", + "issue:28348|issue:43352", + "issue:29870|issue:31515", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:31515|issue:31795", + "issue:44230|issue:45310", + "issue:43298|issue:44230", + "issue:43717|issue:46144", + "issue:42936|issue:46018", + "issue:43668|issue:44031" + ], + "prepared_review_unit_hash": "006cc7982c7a59fe9df36a81f5d08547ea14176b607ed55fba5169ee3b88ff4c", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 39464, + "estimated_input_tokens": 9866, + "estimated_eval_tokens": 19988 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is highly heterogeneous: most soft pairs are only title-similar but concern different models, subsystems, or release regressions. No PRs are present. If a single umbrella issue must be chosen, the broad v5-breakage report is the closest fit, but it is not a true duplicate target for most items.", + "confidence": 0.44, + "canonical_issue_reason": "issue:45850 is the broadest umbrella in the set, describing many model breakages after v5; if one issue must represent the cluster, it is the closest fit.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 is the most representative high-level regression report; the other issues are mostly narrower, unrelated bugs or feature requests.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Different code paths: Qwen3VL timestamping vs GLM-5 RoPE implementation." + }, + { + "left": "issue:45468", + "right": "issue:46894", + "accept": false, + "reason": "Different models and bugs: Gemma audio positional encoding vs CohereASR training-loss shifting." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env CLI vs Qwen3_5 import failure are unrelated." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Saving Pipeline objects and tied-weight-key handling are separate problems." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Timestamp calculation bug is unrelated to inverse_sqrt scheduler kwargs handling." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star cleanup performance issue is unrelated to a VRAM leak." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion about forward args is not the same as a Llama model loading regression." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Different model families and different implementation areas." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to GLM-5 RoPE behavior." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "Shared-expert gating bug and DeepSeek weight-transpose issue are different code paths." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Unexpected network requests from local_files_only is unrelated to Qwen3VL timestamp math." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 API guidance and a dataloader-thread VRAM leak are unrelated." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "Broad v5 regression claim and a specific Qwen3_5 import failure are not the same bug." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamping and processor.save_pretrained missing files are different issues." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return values in GLM-MoE indexing is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is not the same as Gemma audio positional encoding." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Audio continuous batching support and DeepSeek-V4 weight-conversion questions are unrelated." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Qwen3VL timestamp calculation and Qwen3ForCausalLM VRAM leakage are separate bugs." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "Different import failures for different symbols/models; not the same concrete failure." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Nemotron Flash Attention support and Qwen3_5 import failure are unrelated." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both mention tokenizer v5, but one is behavior drift and the other is decoder-type mismatch; not the same code-path bug." + }, + { + "left": "issue:44861", + "right": "issue:47059", + "accept": false, + "reason": "Tied-weight-key crash and CLI completion breakage are unrelated." + }, + { + "left": "issue:28348", + "right": "issue:45250", + "accept": false, + "reason": "GPT2-specific Flash Attention 2.0 support and a generic FA2 request are not the same concrete change." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug is unrelated to the placeholder issue titled 'Unemployment'." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache-strategy typo and an unrelated placeholder issue are not duplicates." + }, + { + "left": "issue:44998", + "right": "issue:45850", + "accept": false, + "reason": "The placeholder issue is unrelated to v5-wide model breakage." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attentions regression is unrelated to the placeholder issue." + }, + { + "left": "issue:44938", + "right": "issue:44998", + "accept": false, + "reason": "Python 3.14 loading failure is unrelated to the placeholder issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight-key crash and assistant tool-call template crash are different bugs." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained loading and an ImportError from configuration_utils are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression is unrelated to tied-weight-key handling." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 Qwen3-VL/Qwen3.5 support request and processor.save_pretrained missing files are different issues." + }, + { + "left": "issue:28348", + "right": "issue:43352", + "accept": false, + "reason": "Both are Flash Attention 2.0 related, but for different models and different concrete support gaps." + }, + { + "left": "issue:29870", + "right": "issue:31515", + "accept": false, + "reason": "BART generate UnboundLocalError and slow checkpoint loading are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLflow reporting on failure and slow from_pretrained loading are unrelated." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint load slowness and tensor-parallel training bugs do not share the same code path." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Loading performance and forward-argument documentation confusion are unrelated." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 inference support and Qwen3.5 from_pretrained loading error are different issues." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Audio continuous batching and Qwen3-VL/Qwen3.5 FP8 support are unrelated." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init_weights behavior and RoFormer attention-interface usage are different problems." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2Former ignore_value handling is unrelated to DeepSeek-V4 shared-expert gating." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "Config type-hint correction and tokenizer regex warning regression are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 26, + "review_unit_index": 62, + "review_unit_count": 62, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-509", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46144" + ], + "soft_pairs": [ + "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": "4c2eb0795ea6e7ffc424e6eb4fa9751ea806f696ded517e3b1b7af2447933ca6", + "component_budget": { + "node_count": 509, + "item_count": 509, + "soft_pair_count": 1871, + "serialized_chars": 970933, + "estimated_input_tokens": 242734, + "estimated_eval_tokens": 485724 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 9, + "serialized_chars": 6638, + "estimated_input_tokens": 1660, + "estimated_eval_tokens": 3576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the issues cover unrelated models, a config type hint, a typo, tokenizer warnings, and a separate attention-interface bug. None of the soft pairs look like the same underlying defect, so they should not be merged as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "If forced to anchor the cluster, issue 43240 is the clearest concrete bug report, but it is not actually duplicated by the others; most other items target different components or different classes of problems.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43240 is the most actionable and specific runtime bug report in the set, with clear symptom language and extra discussion/inbound references, making it the best standalone representative even though the cluster is not a duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Different bugs in different models: Mask2Former ignore_value handling vs. kwargs not forwarded to a loss function." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Unrelated code paths: loss-function kwargs forwarding versus RoFormer attention not using the attention interface." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former loss-parameter usage and tokenizer regex warning are unrelated subsystems and symptoms." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "Different model bugs with no shared concrete code-path; one is loss masking, the other is attention implementation structure." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "A model behavior bug versus a typo in a quantization-related attribute name; not the same underlying issue." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Config type-hint correctness and a spelling/attribute typo are distinct, non-overlapping defects." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Loss-function kwargs forwarding is a runtime behavior bug; ModernBERTConfig norm_eps typing is a separate metadata/type issue." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Llama4 normalization implementation question and ModernBERT config type hint incorrectness are unrelated topics." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Different models and problem types: Mask2Former ignore_value usage versus ModernBERT config typing." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10519-3", + "nodes": [ + "issue:10519", + "issue:11005", + "issue:11307" + ], + "soft_pairs": [ + "issue:10519|issue:11307", + "issue:10519|issue:11005", + "issue:11005|issue:11307" + ], + "prepared_review_unit_hash": "620985b08c09c133bf067c63933fe9b833ae37f17210786825eab26c1a46e41c", + "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 not duplicates: one requests tokenizer truncation-from-start behavior, one asks for ReduceLROnPlateau-like training logic, and one concerns Wav2Vec2 word boundary timestamps. They are separate feature requests in different areas.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue fits this cluster; the items describe unrelated feature requests rather than the same bug or change.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 10519 is the most detailed and clearly scoped of the three, but it is still not a canonical duplicate target because the topics are unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Different requests: tokenizer truncation direction vs Wav2Vec2 word time offsets." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Different functionality: truncation behavior vs ReduceLROnPlateau-style scheduling." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "Unrelated topics: learning-rate scheduling vs speech token boundary timestamps." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 3, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-13244-3", + "nodes": [ + "issue:13244", + "issue:15980", + "issue:16618" + ], + "soft_pairs": [ + "issue:13244|issue:16618", + "issue:15980|issue:16618" + ], + "prepared_review_unit_hash": "16f5ebe336d70c9e95ee6e9061ef2e831443265b3b69ede2aa533065167dfd19", + "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 mostly not duplicates. Two are about model loading, but they point to different failure modes; the Tapas tokenization report is unrelated.", + "confidence": 0.82, + "canonical_issue_reason": "issue:16618 is the broadest and most general model-loading report in the set, so it is the best representative if a single issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:16618 has the clearest generic symptom wording and is the most plausible cluster anchor among the three.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization vs. failing to load bert-base-uncased are different subsystems and different bug classes; no same underlying change or fix is apparent." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Private-model download auth/error-message behavior is not the same as a public model load failure for bert-base-uncased; symptoms and likely code paths differ." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 4, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-26350-2", + "nodes": [ + "issue:26350", + "issue:33260" + ], + "soft_pairs": [ + "issue:26350|issue:33260" + ], + "prepared_review_unit_hash": "78dd7e0ef3c51041f23c21ce5d6299bc07c572339903743979b0cd594d1670c8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1588, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1588, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated community-contribution feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They share only a broad theme of adding architecture support, not the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a duplicate of the other; if one must be kept as a representative issue, there is no clear canonical choice because they request different features and code paths.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different enhancement requests for different subsystems/formats (Flash Attention 2 vs GGUF); they do not describe the same bug or change and would not plausibly be merged into one issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 5, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-9", + "nodes": [ + "issue:27088", + "issue:28936", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370", + "issue:46222", + "issue:46226", + "issue:46931" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:27088|issue:38435", + "issue:27088|issue:28936", + "issue:28936|issue:43834", + "issue:28936|issue:44365", + "issue:28936|issue:38435", + "issue:28936|issue:44370", + "issue:38435|issue:44370", + "issue:38435|issue:44365", + "issue:38435|issue:43834", + "issue:38435|issue:46222", + "issue:38435|issue:46226", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:43834|issue:46222", + "issue:43834|issue:46226", + "issue:43834|issue:46931", + "issue:44365|issue:44370", + "issue:44365|issue:46222", + "issue:44365|issue:46226", + "issue:44365|issue:46931", + "issue:44370|issue:46222", + "issue:44370|issue:46226", + "issue:44370|issue:46931", + "issue:46222|issue:46226", + "issue:46222|issue:46931", + "issue:46226|issue:46931" + ], + "prepared_review_unit_hash": "7d8c929d9a4eaa5c25318a5855fd2f20d73dd423eb609764c8045d86633165b7", + "component_budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 29, + "serialized_chars": 15980, + "estimated_input_tokens": 3995, + "estimated_eval_tokens": 8246 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 29, + "serialized_chars": 15980, + "estimated_input_tokens": 3995, + "estimated_eval_tokens": 8246 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mixed. The visible metadata supports one exact duplicate pair, 46222/46226, with identical Thai titles and matching body metadata. The other issues split into distinct explicit i18n titles ([i18n-TR], [i18n-es], [i18n-ro]) and several placeholder-title issues with no body/link metadata strong enough to merge on title similarity alone.", + "confidence": 0.87, + "canonical_issue_reason": "46222 is the best canonical issue for the only clearly duplicated subgroup because it is an exact match with 46226 and has inbound references.", + "canonical_pr_reason": null, + "best_issue_reason": "46222 is the strongest overall issue anchor for triage because it represents the only exact duplicate pair in the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Different visible titles: a specific Turkish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Different visible language-coded titles ([i18n-TR] vs [i18n-ro]); the metadata shows separate translation requests." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Different visible language-coded titles ([i18n-TR] vs [i18n-es]); the metadata shows separate translation requests." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Different visible language-coded titles ([i18n-es] vs [i18n-ro]); the metadata shows separate translation requests." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Different visible titles: a specific Spanish i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Different visible titles: a specific Romanian i18n request versus a placeholder translation template; no body or link metadata ties them together." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "The visible titles are unrelated: a Romanian i18n request versus a Thai-titled issue; the packet shows no shared concrete request." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "The visible titles are unrelated: a Romanian i18n request versus a Thai-titled issue; the packet shows no shared concrete request." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:43834", + "right": "issue:46931", + "accept": false, + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "Same placeholder title, but the packet has no body text, explicit targets, or references showing the same concrete issue." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:44365", + "right": "issue:46931", + "accept": false, + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "The visible titles do not identify the same request: placeholder translation template versus a Thai-titled issue, with no shared metadata." + }, + { + "left": "issue:44370", + "right": "issue:46931", + "accept": false, + "reason": "Different visible titles and no shared body/link metadata; the placeholder template does not match the Thai title." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Exact duplicate: the Thai titles and visible body metadata match, and 46222 has inbound references." + }, + { + "left": "issue:46222", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai titles with different visible text, so they are not the same issue." + }, + { + "left": "issue:46226", + "right": "issue:46931", + "accept": false, + "reason": "Different Thai titles with different visible text, so they are not the same issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The exact duplicate call on 46222/46226 is grounded, but the broader summary and several soft-edge reasons go beyond the packet. In particular, 43834/44365/44370 all have the same placeholder title and only soft-similarity evidence; the packet does not show they are different language requests, so that part is overconfident. Please keep the cluster description and pair reasons strictly tied to the visible titles/body metadata." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36718-21", + "nodes": [ + "issue:36718", + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:42954", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133", + "issue:46645", + "issue:46653", + "issue:47004" + ], + "soft_pairs": [ + "issue:36718|issue:47004", + "issue:36718|issue:41115", + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36817|issue:42954", + "issue:36879|issue:41211", + "issue:36879|issue:43909", + "issue:36879|issue:44121", + "issue:36879|issue:43837", + "issue:36879|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:36879|issue:37015", + "issue:37015|issue:42503", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42954", + "issue:42738|issue:45027", + "issue:42503|issue:42738", + "issue:42738|issue:42954", + "issue:42738|issue:46653", + "issue:42954|issue:44205", + "issue:42954|issue:43671", + "issue:42954|issue:45177", + "issue:42954|issue:46653", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43837|issue:44121", + "issue:43909|issue:44205", + "issue:43909|issue:46645", + "issue:44121|issue:45322", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133", + "issue:42503|issue:46653", + "issue:45027|issue:46653" + ], + "prepared_review_unit_hash": "260743dceed40fee5e79a496bd50cd62850da40cb93f0ee662395e8db5c37bfe", + "component_budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28621, + "estimated_input_tokens": 7156, + "estimated_eval_tokens": 14568 + }, + "budget": { + "node_count": 21, + "item_count": 21, + "soft_pair_count": 48, + "serialized_chars": 28621, + "estimated_input_tokens": 7156, + "estimated_eval_tokens": 14568 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is only loosely related by topic (mostly model-addition/support requests); the items are not duplicate-level matches and span different models, modalities, and use cases. No PRs are present.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43837 is the most central concrete support request in the similarity graph, but this is not a true duplicate cluster; the other items are mostly distinct model requests.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43837 is the best single anchor only because it is a clear, specific model-support request with many soft links; however, the linked issues are still different changes, so it should not be treated as a duplicate canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36718", + "right": "issue:47004", + "accept": false, + "reason": "Both are keypoint matcher requests, but RoMa vs LoMa are different models and changes; not the same underlying feature." + }, + { + "left": "issue:36718", + "right": "issue:41115", + "accept": false, + "reason": "Keypoint matching vs MiniCPM3 architecture are unrelated model requests." + }, + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "EuroBert config addition and DEIMv2 support are different model entries." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config work and Qwen3-TTS support are different model families and modalities." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert and RF-DETR are unrelated models; same broad area only." + }, + { + "left": "issue:36817", + "right": "issue:42954", + "accept": false, + "reason": "EuroBert config and XiaoMi MiMo are different model support requests." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are different detection models, not one concrete change." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR and LFM2.5 Audio 1.5B are unrelated model additions." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR vs a weight-sparse transformer request are different code paths and model types." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR and Qwen3-ASR are different models and different support work." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR and Qwen3-TTS are unrelated features." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different model additions." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT and the weight-sparse transformer request are different architectures and tasks." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio 1.5B are unrelated requests." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are different models; no shared concrete bug/change." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are different model additions." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture and Qwen3-ASR support are different model families and modalities." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 architecture and the weight-sparse transformer request are unrelated changes." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 architecture and ModernVBERT models are different requests." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 architecture and Voxtral TTS support are distinct model tasks." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are different model support requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT models and Qwen3-ASR support are different model families and tasks." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT models and Qwen3-TTS support are unrelated." + }, + { + "left": "issue:42503", + "right": "issue:42954", + "accept": false, + "reason": "ModernVBERT and XiaoMi MiMo are different model additions." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like models with RoPE and Voxtral TTS support are different feature requests." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT models and BERT-like RoPE support are overlapping only at a broad BERT theme, not the same change." + }, + { + "left": "issue:42738", + "right": "issue:42954", + "accept": false, + "reason": "RoPE support for BERT-like models and XiaoMi MiMo are different 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": "XiaoMi MiMo and SAM3-LiteText are different model requests." + }, + { + "left": "issue:42954", + "right": "issue:43671", + "accept": false, + "reason": "XiaoMi MiMo and Qwen3-TTS are different models and modalities." + }, + { + "left": "issue:42954", + "right": "issue:45177", + "accept": false, + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are unrelated model additions." + }, + { + "left": "issue:42954", + "right": "issue:46653", + "accept": false, + "reason": "XiaoMi MiMo and Mirko Privitera are unrelated." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are related only by naming; they are different modalities and separate support changes." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral TTS are different model integrations, not the same change." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model requests." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support and the weight-sparse transformer request are different features." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio 1.5B are both audio-related but are different models and different support work." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE are different model requests." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support and the weight-sparse transformer request are unrelated." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are different audio-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 are different support requests." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are different model requests despite both being vision-language related." + }, + { + "left": "issue:42503", + "right": "issue:46653", + "accept": false, + "reason": "ModernVBERT models and Mirko Privitera are unrelated." + }, + { + "left": "issue:45027", + "right": "issue:46653", + "accept": false, + "reason": "Voxtral TTS support and Mirko Privitera 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": 7, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-38740-2", + "nodes": [ + "issue:38740", + "issue:46443" + ], + "soft_pairs": [ + "issue:38740|issue:46443" + ], + "prepared_review_unit_hash": "e6efc02b755ec3068c4726697eb6e754dabb330486ee94fb32a1be05cd6df1b6", + "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 issues are unrelated: one is a docs request to add Pruna as an optimization framework, the other is a Vision RoPE refactor. The soft similarity signal is too weak to indicate duplication.", + "confidence": 0.98, + "canonical_issue_reason": "If a single issue must represent the cluster, issue 46443 is the more substantive engineering task; issue 38740 is a documentation-oriented request and is clearly not the same problem.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46443 is the better representative because it describes an actual code refactor in a specific model component, whereas issue 38740 is a standalone docs enhancement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Reject: these concerns do not describe the same underlying bug or change. One is a docs addition about Pruna, the other is a Vision RoPE refactor in a different subsystem with no shared targets or filenames." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 8, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41084-2", + "nodes": [ + "issue:41084", + "issue:43037" + ], + "soft_pairs": [ + "issue:41084|issue:43037" + ], + "prepared_review_unit_hash": "31e8894881b851c195a93399ab05d1b027afea12e41c25f5d1afe757e28008f5", + "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 about different topics: one requests block decoding behavior, while the other asks for DeepSeek v3.2 model support. There is no sign they describe the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": "No canonical issue: the titles and implied scopes are unrelated, with no shared targets or code-level evidence tying them together.", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good global best for the cluster because they do not form a duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "These are different feature requests/problems: block decoding vs model-version support. The similarity is only superficial and does not indicate the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 9, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41306-2", + "nodes": [ + "issue:41306", + "issue:41826" + ], + "soft_pairs": [ + "issue:41306|issue:41826" + ], + "prepared_review_unit_hash": "fe7363c30fa1660cb10a7f05ea4dc3e075e2428e2135f716f0bc8b717edf3dde", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1585, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1585, + "estimated_input_tokens": 397, + "estimated_eval_tokens": 1050 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues both mention tiled/memory-saving model components, but they target different features and code paths: FLCE/LigerFusedLinearCrossEntropy flag wiring in `from_pretrained()` versus integrating TiledMLP for lower memory use. They do not look like the same underlying bug or change request.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 41826 is the broader feature request around memory-footprint reduction; 41306 is a narrower configuration/wiring request for a different component.", + "canonical_pr_reason": null, + "best_issue_reason": "41826 is the better cluster representative because it is the more general and product-facing request about memory savings, while 41306 is a specific flag exposure problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41306", + "right": "issue:41826", + "accept": false, + "reason": "Both are about tiled optimization/memory reduction, but one is specifically about exposing a FLCE flag in `from_pretrained()` and the other about adding TiledMLP support. Different components and likely different implementation paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43598-2", + "nodes": [ + "issue:43598", + "issue:43600" + ], + "soft_pairs": [ + "issue:43598|issue:43600" + ], + "prepared_review_unit_hash": "9f6090a3384c2bdef18d1f96996cb6c2a220e8ffde747d5725676de80386656b", + "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 both about custom trainers in `Trainer`, but they describe different problems: one is about `remove_unused_column` customizability, the other about FLOPS estimation. They are related in theme but not duplicates.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #43598 is the broader, more representative Trainer customizability concern; #43600 is a separate FLOPS-estimation problem.", + "canonical_pr_reason": null, + "best_issue_reason": "#43598 is the better cluster representative because it targets the core Trainer customization behavior rather than a narrower accounting edge case.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "They share the general context of custom trainers, but the underlying problems differ: one concerns `remove_unused_column`, the other concerns FLOPS estimation. 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": 11, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43684-2", + "nodes": [ + "issue:43684", + "issue:44405" + ], + "soft_pairs": [ + "issue:43684|issue:44405" + ], + "prepared_review_unit_hash": "ca396277310ed4044d4b6a6c3b1d9494e027e32a1429e88c6bd5499e0e98e106", + "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 both model-registration requests, but they target different model families and different auto classes, so they do not look like the same underlying issue.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43684 is the broader and earlier request: it asks for Qwen3-Omni registration in both AutoModel and AutoModelForConditionalGeneration, which makes it the better representative of the Qwen3-Omni registration gap.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43684 is the best issue candidate because it covers a concrete registration omission and is broader in scope than the sequence-classification-only request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different targets: Qwen3-Omni auto-model registration versus Qwen3.5 AutoModelForSequenceClassification support. They concern different model variants and different auto-class paths, so they are 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": 12, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43835-4", + "nodes": [ + "issue:43835", + "issue:43845", + "issue:43846", + "issue:43859" + ], + "soft_pairs": [ + "issue:43835|issue:43845", + "issue:43835|issue:43846", + "issue:43835|issue:43859", + "issue:43845|issue:43846", + "issue:43845|issue:43859", + "issue:43846|issue:43859" + ], + "prepared_review_unit_hash": "070f5b608a857c15247f2e59059837d7dbad9a05a4536d392148dbeaa54bfe0a", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The duplicate cluster is most plausibly the three closed issues titled \"huggingface\"; 43835 titled \"Fett\" looks separate. The evidence here is limited to title matching and soft-similarity scores, so the duplicate call should stay conservative.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 43845 is the earliest of the three matching \"huggingface\" reports and is a reasonable representative for that duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43845 is the best issue representative because it is the earliest item in the likely duplicate trio and matches the shared title pattern.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles; with only soft similarity and no shared targets/files, 43835 ('Fett') does not look like the same report as the \"huggingface\" issues." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no concrete overlap beyond soft similarity, so this does not look like the same underlying issue." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "\"Fett\" is distinct from the repeated \"huggingface\" title, and there is no stronger evidence of a shared bug report." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "These have the same title and a high soft-similarity score, so they plausibly refer to the same underlying issue." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "These share the same title and show strong soft similarity, which is enough to treat them as likely duplicates." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Same title plus strong soft-similarity evidence makes these look like the same underlying issue report." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The clustering summary is mostly grounded, but the analyst overstates the evidence in a few places. In particular, the packet only shows matching titles and soft-similarity scores; it does not provide body text, so phrases like \u201crepeated title/body pattern\u201d and \u201cnear-identical\u201d are stronger than the evidence supports. The duplicate judgment among the three \u2018huggingface\u2019 issues is plausible, but the rationale should stay closer to the available data and be more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44371-6", + "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": "0491817e9d6becd235c32fe98290b96df50e972ed17cfc49d13c3e22df575429", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7355, + "estimated_input_tokens": 1839, + "estimated_eval_tokens": 3934 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7355, + "estimated_input_tokens": 1839, + "estimated_eval_tokens": 3934 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All items look like unrelated spam/placeholder issues rather than duplicates of one underlying bug or change. No PRs are present, and none of the soft-similarity pairs should be merged.", + "confidence": 0.97, + "canonical_issue_reason": "No canonical issue stands out because the items do not describe the same problem; they are mostly spammy or generic posts with unrelated titles.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a good representative for the cluster because there is no shared technical topic to consolidate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are generic spam placeholders, but they do not describe the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Both are generic spam placeholders, but they are not the same bug/change." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "One is spam and the other is a vague documentation suggestion; no shared underlying problem." + }, + { + "left": "issue:44371", + "right": "issue:47036", + "accept": false, + "reason": "Unrelated spam placeholder versus community tip; not a duplicate report." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "Different unrelated spam content; no common bug or change." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "A vague documentation idea is not the same as a spam placeholder issue." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "These do not describe the same concrete problem or change." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Both are spam placeholders, but they are distinct unrelated posts rather than one report." + }, + { + "left": "issue:44572", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder versus benchmark tip; no shared underlying issue." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Different unrelated spam content, not a duplicate." + }, + { + "left": "issue:44573", + "right": "issue:47036", + "accept": false, + "reason": "These are unrelated posts and do not target the same problem." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "Different content and intent; not the same issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45145-2", + "nodes": [ + "issue:45145", + "issue:45151" + ], + "soft_pairs": [ + "issue:45145|issue:45151" + ], + "prepared_review_unit_hash": "1fa3478aa249d6401b7780a4d0bccd30a0030a90cc6ad66e4116c3eb9c341985", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are likely duplicates: they share the exact title, were opened within hours of each other, have very similar body lengths, and the pair has a soft_similarity signal. #45145 is the best representative because it is earlier and has the only inbound reference.", + "confidence": 0.79, + "canonical_issue_reason": "Choose #45145 as canonical because it is the earlier filing and has slightly more link signal (one inbound reference), making it the better representative for this likely duplicate report.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the best issue representative because it was opened first and has the stronger external signal among the two; #45151 appears to be a later duplicate of the same report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "Same title, close creation times, similar body lengths, and a soft_similarity signal make this likely a duplicate issue pair, even though the evidence is not strong enough to be definitive." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The conclusion is plausible, but the analyst overstates the evidence. The packet only directly supports same title, close creation times, similar body length, and a soft_similarity signal; it does not show the bodies are actually 'strongly similar' or that there is 'no distinguishing metadata.' The 0.98 confidence is also too strong for this level of evidence. The soft-edge accept=true is a bit too assertive given deterministic_accept=false and only moderate similarity signal." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45295-2", + "nodes": [ + "issue:45295", + "issue:45373" + ], + "soft_pairs": [], + "prepared_review_unit_hash": "f5fd95fada2f530ab0999e343c17c16266cde0c7182bb22509972ca44827cd49", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues describe the same missing Gemma 4 sequence-classification support; the later, more specific open issue is the better cluster representative.", + "confidence": 0.94, + "canonical_issue_reason": "issue #45373 is the clearest canonical tracker: it is open, explicitly names the missing class `Gemma4ForSequenceClassification`, and states the module-level gap relative to Gemma 2/3.", + "canonical_pr_reason": null, + "best_issue_reason": "issue #45373 best captures the concrete request and likely remaining work, while #45295 is a shorter earlier duplicate with less detail and is already closed.", + "best_pr_reason": null, + "soft_edge_verdicts": [] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 16, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45762-2", + "nodes": [ + "issue:45762", + "issue:45763" + ], + "soft_pairs": [ + "issue:45762|issue:45763" + ], + "prepared_review_unit_hash": "ffe7623028f1e98b327a55a36c73f25b2c3ed9ba971e0774de75e730655ce96e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1476, + "estimated_input_tokens": 369, + "estimated_eval_tokens": 994 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1476, + "estimated_input_tokens": 369, + "estimated_eval_tokens": 994 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two closed issues look like unrelated spam/personal-name posts, not duplicates of the same bug report. There is no clear canonical issue in this cluster.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a meaningful bug report or a better representative of a shared problem; both appear to be standalone spam/irrelevant entries.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "The titles are different personal names with different dates and contain no shared bug context, target, or description indicating the same underlying issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 17, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45764-2", + "nodes": [ + "issue:45764", + "issue:45765" + ], + "soft_pairs": [ + "issue:45764|issue:45765" + ], + "prepared_review_unit_hash": "137225402133f273d8829f6743d80b0fc649d8160a718b1254a8e4b09f8b658f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues are unrelated, name/date-style posts with no technical content or shared problem statement, so they should not 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:45764", + "right": "issue:45765", + "accept": false, + "reason": "Different named entries with different dates and no body/context indicating the same underlying issue; this does not look like 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": 18, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45810-2", + "nodes": [ + "issue:45810", + "issue:46295" + ], + "soft_pairs": [ + "issue:45810|issue:46295" + ], + "prepared_review_unit_hash": "27b884051499e6204dc95cf09f9a35183a96f22dbafece899c6fba8d11845b72", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are not duplicates: one requests a new Qwen3_5 token-classification model class, while the other requests FP8 training support. They touch different features and code paths.", + "confidence": 0.98, + "canonical_issue_reason": "issue:46295 is the broader, more generally scoped request and has more inbound discussion, so it is the better representative issue in this set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46295 is the most central standalone issue here because it appears broader and more active than the model-addition request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different underlying requests: adding a token-classification model class vs enabling FP8 training support. They do not describe 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": 19, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45869-5", + "nodes": [ + "issue:45869", + "issue:45870", + "issue:45871", + "issue:45872", + "issue:45873" + ], + "soft_pairs": [ + "issue:45869|issue:45870", + "issue:45869|issue:45871", + "issue:45870|issue:45871", + "issue:45870|issue:45872", + "issue:45870|issue:45873", + "issue:45871|issue:45872", + "issue:45871|issue:45873", + "issue:45872|issue:45873" + ], + "prepared_review_unit_hash": "64806193afc2b183845d500e1f23ea840e9531fd4d2a51e692e972b92dae658c", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5724, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5724, + "estimated_input_tokens": 1431, + "estimated_eval_tokens": 3118 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "There are two clear duplicate issue pairs: 45870/45871 and 45872/45873. The dataset-bundle issue (45869) is only loosely related to the others and should not be merged with them.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 45870 is a good canonical representative because it is the earliest exact duplicate pair for the repeated \u201cExample: Financial time series prediction with DataForge\u201d report and is cleaner to retain as the primary thread.", + "canonical_pr_reason": null, + "best_issue_reason": "45870 is the best single issue to keep as the representative item among the near-identical reports; it has an exact duplicate and the title is the clearest standalone example.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Same general finance/dataset theme, but this looks like a dataset offering versus an example post; not the same underlying issue." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Related topic, but the dataset-bundle post is not the same concrete report as the example issue." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Exact title match and same body shape; these are the same issue." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both mention DataForge and finance, but one is an example and the other is a scraping-service post; different underlying requests." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Same as 45870/45872: only broadly similar theme, not the same concrete issue." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Exact duplicate of 45870, but not the same underlying issue as the scraping-service request." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Same reason as 45871/45872; different request despite shared finance/DataForge wording." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Exact title match and same body shape; these are the same issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: the identical-title, jaccard-1.0 pairs 45870/45871 and 45872/45873 are reasonably treated as duplicate clusters, while 45869 is only loosely related and kept separate. The soft-edge verdicts are conservative and align with the available evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46439-2", + "nodes": [ + "issue:46439", + "issue:46441" + ], + "soft_pairs": [ + "issue:46439|issue:46441" + ], + "prepared_review_unit_hash": "79b1e5f076b6b55be1de1645a7e0b60bc71d0dc1c6aaa634d9bf2741d5bdab4e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both issues concern `cache_implementation=\"static\"` and `linear_attention`, but they describe different failure modes: one is a shape-corruption `RuntimeError`, the other a `KeyError` in Qwen3.5. Related, but not clearly the same underlying bug from the available evidence.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it describes the broader root-cause class (`early_initialization` breaking linear-attention layers) rather than a single model-specific crash symptom.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46439 is the best issue candidate overall: it is more general, points at the likely underlying static-cache initialization problem, and can subsume narrower reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46439", + "right": "issue:46441", + "accept": false, + "reason": "Same area, but the reported bugs differ concretely (`RuntimeError` shape corruption vs `KeyError: 'linear_attention'`), so there is not enough evidence they are 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": 21, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46551-2", + "nodes": [ + "issue:46551", + "issue:46552" + ], + "soft_pairs": [ + "issue:46551|issue:46552" + ], + "prepared_review_unit_hash": "66c3e2d38afcea752b232ce89d261d55dac4eac22e82b348091e36dd403f0ce9", + "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": "The two issues are exact duplicates: same title, same body, same bug in Wav2Vec2CTCTokenizer v5. Issue #46552 is the better canonical issue because it is open and already has inbound references, while #46551 is closed.", + "confidence": 0.99, + "canonical_issue_reason": "#46552 is the stronger canonical issue: it is open, has inbound references, and matches the same exact report as #46551.", + "canonical_pr_reason": null, + "best_issue_reason": "#46552 best represents the cluster because it is the active report for the identical tokenizer special-token leak bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46551", + "right": "issue:46552", + "accept": true, + "reason": "Exact duplicate issue reports: identical title, identical body, and the same underlying Wav2Vec2CTCTokenizer v5 special-token bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well-grounded in the packet: both issues have the same title, the same body length, and a Jaccard similarity of 1.0, so treating them as duplicates is reasonable. The canonical choice of #46552 is also supported by the packet because it is open and has inbound references, while #46551 is closed. No mergeability issues apply since this is an issue pair, not a PR pair." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 22, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46657-2", + "nodes": [ + "issue:46657", + "issue:46692" + ], + "soft_pairs": [ + "issue:46657|issue:46692" + ], + "prepared_review_unit_hash": "fde121ff9bf08ef0ccbb105bb260e155feedd88c5b7f3405d3836f6251ddfa9c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1568, + "estimated_input_tokens": 392, + "estimated_eval_tokens": 1040 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1568, + "estimated_input_tokens": 392, + "estimated_eval_tokens": 1040 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are in the same broad training/object-detection area, but they describe different problems: a tensor shape mismatch versus an out-of-memory condition after several epochs. They should not be deduplicated.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 46657 is the more specific, concrete report and best anchor for the cluster; 46692 describes a different failure mode (OOM during RT-DETR v2 training) rather than the same underlying bug.", + "canonical_pr_reason": null, + "best_issue_reason": "46657 is the better representative issue because it names a precise runtime error ('size of tensors mismatch'). 46692 appears to be a separate resource/memory issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46657", + "right": "issue:46692", + "accept": false, + "reason": "Reject: the titles point to different underlying bugs and symptoms. A tensor-size mismatch during object-detection training is not the same as memory exhaustion after some epochs in RT-DETR v2, even though both occur in training." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 23, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46822-2", + "nodes": [ + "issue:46822", + "issue:46823" + ], + "soft_pairs": [ + "issue:46822|issue:46823" + ], + "prepared_review_unit_hash": "ab1dfb8e4df510766853854e845d4e332838cdd4bfc22e1323e896fc929d2408", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two different loading bugs in different code paths: ZeRO-3 zero.Init partitioning/OOM vs 4-bit device_map/max_memory CPU dispatch. They should not be treated as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue; the two issues describe distinct failure modes in separate subsystems (ZeRO-3 partitioning vs 4-bit auto-dispatch behavior).", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good cluster representative for the other because they are unrelated bugs and not mergeable into a single duplicate target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46822", + "right": "issue:46823", + "accept": false, + "reason": "Different underlying bugs: one is ZeRO-3 zero.Init failing to partition a composite submodule, the other is 4-bit loading with device_map/max_memory sending modules to CPU incorrectly." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 24, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46906-9", + "nodes": [ + "issue:46906", + "issue:46940", + "issue:46948", + "issue:46969", + "issue:46985", + "issue:46998", + "issue:47037", + "issue:47049", + "issue:47053" + ], + "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:46998", + "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:47037|issue:47049", + "issue:47037|issue:47053", + "issue:47049|issue:47053", + "issue:46969|issue:47049", + "issue:46985|issue:47053", + "issue:46969|issue:47053" + ], + "prepared_review_unit_hash": "6284bb4510c0fb40caacb43971c8f48d3f4a2e31deac9bd6cecca40555e357d1", + "component_budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 27, + "serialized_chars": 14785, + "estimated_input_tokens": 3697, + "estimated_eval_tokens": 7650 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 27, + "serialized_chars": 14785, + "estimated_input_tokens": 3697, + "estimated_eval_tokens": 7650 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These look like a dated series of daily triage issues for the same area, but the available metadata does not show a shared underlying bug or fix. I would not merge them as duplicates on title/template similarity alone.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 46906 is the earliest in the series and has the strongest inbound reference count, so it is the best anchor if the cluster is being represented by a single issue.", + "canonical_pr_reason": null, + "best_issue_reason": "Among the issues, 46906 is the most representative hub: earliest created and most-linked, with no evidence that another issue is a better canonical representative.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46906", + "right": "issue:46940", + "accept": false, + "reason": "Same triage template and adjacent dates, but no evidence they track the same underlying failure; likely separate daily reports." + }, + { + "left": "issue:46906", + "right": "issue:46948", + "accept": false, + "reason": "Strong textual similarity, but these read as distinct day-specific triage entries rather than one duplicate defect." + }, + { + "left": "issue:46906", + "right": "issue:46969", + "accept": false, + "reason": "Similar series title only; no body or diff evidence that the underlying incident is the same." + }, + { + "left": "issue:46906", + "right": "issue:46985", + "accept": false, + "reason": "Date-stamped triage posts are not enough to establish duplicate status; likely separate reports." + }, + { + "left": "issue:46906", + "right": "issue:47037", + "accept": false, + "reason": "Both are triage issues, but the metadata only supports a recurring template, not one shared bug." + }, + { + "left": "issue:46940", + "right": "issue:46948", + "accept": false, + "reason": "Adjacent triage posts with the same template, but no concrete evidence of the same failure mode." + }, + { + "left": "issue:46940", + "right": "issue:46969", + "accept": false, + "reason": "Looks like another daily triage entry in the series, not a duplicate of a specific underlying issue." + }, + { + "left": "issue:46940", + "right": "issue:46985", + "accept": false, + "reason": "High similarity is explained by the repeated triage format; no shared bug details are available." + }, + { + "left": "issue:46940", + "right": "issue:47049", + "accept": false, + "reason": "Same broad subsystem and template, but not enough to say these are the same incident." + }, + { + "left": "issue:46948", + "right": "issue:46969", + "accept": false, + "reason": "Similar dated triage entries, but nothing indicates the same root cause or resolution target." + }, + { + "left": "issue:46948", + "right": "issue:46985", + "accept": false, + "reason": "These appear to be separate daily reports rather than duplicate issues." + }, + { + "left": "issue:46948", + "right": "issue:46998", + "accept": false, + "reason": "Template similarity only; no evidence of one underlying bug being tracked twice." + }, + { + "left": "issue:46969", + "right": "issue:46985", + "accept": false, + "reason": "Same triage-series pattern, but not enough signal to merge as duplicates." + }, + { + "left": "issue:46969", + "right": "issue:46998", + "accept": false, + "reason": "The titles match a recurring report format; no concrete shared failure is visible." + }, + { + "left": "issue:46969", + "right": "issue:47037", + "accept": false, + "reason": "Likely separate daily triage items; similarity is too generic to prove duplication." + }, + { + "left": "issue:46969", + "right": "issue:47049", + "accept": false, + "reason": "No evidence beyond the repeated triage naming convention that these describe the same bug." + }, + { + "left": "issue:46969", + "right": "issue:47053", + "accept": false, + "reason": "Same series, but the available metadata does not establish a single underlying defect." + }, + { + "left": "issue:46985", + "right": "issue:46998", + "accept": false, + "reason": "Two consecutive triage reports; similar wording is expected and does not imply duplicate status." + }, + { + "left": "issue:46985", + "right": "issue:47037", + "accept": false, + "reason": "No body/diff evidence of the same code-path problem, only the shared triage template." + }, + { + "left": "issue:46985", + "right": "issue:47049", + "accept": false, + "reason": "These look like separate daily issue reports, not one issue reopened under a new date." + }, + { + "left": "issue:46985", + "right": "issue:47053", + "accept": false, + "reason": "Insufficient evidence of the same underlying failure; the match is only in the recurring title format." + }, + { + "left": "issue:46998", + "right": "issue:47037", + "accept": false, + "reason": "Recurring integration-triage formatting is not enough to treat these as duplicates." + }, + { + "left": "issue:46998", + "right": "issue:47049", + "accept": false, + "reason": "The metadata suggests a series of reports, not one repeated bug report." + }, + { + "left": "issue:46998", + "right": "issue:47053", + "accept": false, + "reason": "Same template, different day; no concrete shared defect evidence." + }, + { + "left": "issue:47037", + "right": "issue:47049", + "accept": false, + "reason": "Similar titles alone do not show the same underlying failure or fix." + }, + { + "left": "issue:47037", + "right": "issue:47053", + "accept": false, + "reason": "Likely separate triage entries in a daily sequence rather than duplicates." + }, + { + "left": "issue:47049", + "right": "issue:47053", + "accept": false, + "reason": "Adjacent date-stamped triage reports; no indication they are the same issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 25, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46983-2", + "nodes": [ + "issue:46983", + "issue:46984" + ], + "soft_pairs": [ + "issue:46983|issue:46984" + ], + "prepared_review_unit_hash": "f78da9fb09e1e776166274b0a1415aa095bfdd39e6672e232ac37d3ae2ee14d4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1490, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1490, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These look like likely duplicates of the same RWKV7 support request: identical titles, very close creation times, and a strong similarity signal. The evidence is limited because the packet does not expose body text or linking metadata, so the conclusion should be treated as provisional rather than certain.", + "confidence": 0.74, + "canonical_issue_reason": "#46984 is the better canonical issue thread because it is open and appears to be the surviving discussion for the same request, but this is based mainly on the identical title, near-simultaneous creation time, and soft similarity signal since the body content is not visible.", + "canonical_pr_reason": null, + "best_issue_reason": "#46984 is the best issue to keep because it is open and likely represents the active thread for this same RWKV7 support request; the available evidence is strong enough for a duplicate triage, but not enough to claim full certainty.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46983", + "right": "issue:46984", + "accept": true, + "reason": "Same title, created minutes apart, and a strong similarity score make them likely the same underlying request. There is no body text or cross-link evidence in the packet, so this is a cautious accept rather than a definitive one." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The duplicate/canonical choice is plausible, but the writeup is a bit overconfident for the evidence available. The packet only shows same title, close timestamps, and a soft similarity signal; it does not expose body content, and there is a meaningful metadata difference (closed vs open, plus different discussion activity). Please phrase the conclusion more conservatively and lower the confidence unless you can cite stronger evidence from the bodies or links." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 26, + "component_count": 26, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-47007-2", + "nodes": [ + "issue:47007", + "issue:47015" + ], + "soft_pairs": [ + "issue:47007|issue:47015" + ], + "prepared_review_unit_hash": "b7bfdf0af3640fe8affd04e43152543eb066c5432b84ab2ba4a10b3fc46e661d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1625, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1625, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are broadly about AI-assisted cybersecurity vulnerability discovery, but they read like distinct writeups rather than the same underlying issue. I would not merge them as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "Issue 47015 is the broader, more general formulation of the topic, so it is the better representative if a single issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47015 is the most general and concise description of the shared theme, but the pair does not look like a true duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47007", + "right": "issue:47015", + "accept": false, + "reason": "The titles overlap on AI-driven vulnerability discovery, but they describe different angles ('lessons from using advanced LLMs' vs 'novel application of transformers') and there is no evidence they refer to the same underlying report or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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-40546-48", + "nodes": [ + "pull_request:43451", + "pull_request:45694", + "pull_request:46196", + "pull_request:46266", + "pull_request:46658", + "pull_request:46707", + "pull_request:46836", + "pull_request:46903", + "pull_request:46907", + "pull_request:46908", + "pull_request:46912", + "pull_request:46926", + "pull_request:46927", + "pull_request:46928", + "pull_request:46941", + "pull_request:46949", + "pull_request:46950", + "pull_request:46951", + "pull_request:46954", + "pull_request:46957", + "pull_request:46958", + "pull_request:46961", + "pull_request:46965", + "pull_request:46968", + "pull_request:46970", + "pull_request:46971", + "pull_request:46974", + "pull_request:46986", + "pull_request:46995", + "pull_request:47003", + "pull_request:47006", + "pull_request:47019", + "pull_request:47025", + "pull_request:47034", + "pull_request:47035", + "pull_request:47054", + "pull_request:47055", + "pull_request:47056", + "pull_request:47057", + "pull_request:47058", + "pull_request:47062", + "pull_request:47069", + "pull_request:47077", + "pull_request:47078", + "pull_request:47079" + ], + "soft_pairs": [ + "pull_request:46903|pull_request:46958", + "pull_request:46950|pull_request:46951", + "pull_request:46950|pull_request:47057", + "pull_request:46927|pull_request:47054", + "pull_request:46927|pull_request:46951", + "pull_request:46951|pull_request:46971", + "pull_request:46949|pull_request:46971", + "pull_request:46951|pull_request:47057", + "pull_request:46941|pull_request:46970", + "pull_request:47056|pull_request:47058", + "pull_request:47006|pull_request:47058", + "pull_request:46927|pull_request:47055", + "pull_request:46908|pull_request:47006", + "pull_request:46908|pull_request:46928", + "pull_request:46907|pull_request:46926", + "pull_request:46927|pull_request:46986", + "pull_request:46986|pull_request:47058", + "pull_request:46926|pull_request:46928", + "pull_request:46928|pull_request:47057", + "pull_request:46926|pull_request:46941", + "pull_request:46908|pull_request:47058", + "pull_request:46928|pull_request:47058", + "pull_request:46928|pull_request:47054", + "pull_request:46941|pull_request:46986", + "pull_request:46941|pull_request:47006", + "pull_request:46954|pull_request:46961", + "pull_request:46928|pull_request:46941", + "pull_request:46958|pull_request:46961", + "pull_request:46954|pull_request:46968", + "pull_request:46954|pull_request:46958", + "pull_request:46961|pull_request:46974", + "pull_request:46958|pull_request:46974", + "pull_request:46954|pull_request:46971", + "pull_request:46971|pull_request:46974", + "pull_request:46958|pull_request:46968", + "pull_request:46968|pull_request:46974", + "pull_request:46961|pull_request:46968", + "pull_request:46961|pull_request:46971", + "pull_request:46949|pull_request:46968", + "pull_request:47077|pull_request:47078", + "pull_request:46266|pull_request:46957", + "pull_request:47019|pull_request:47025", + "pull_request:47062|pull_request:47077", + "pull_request:47034|pull_request:47035", + "pull_request:47019|pull_request:47069", + "pull_request:46196|pull_request:46995", + "pull_request:46957|pull_request:46965", + "pull_request:46266|pull_request:46707", + "pull_request:46707|pull_request:46961", + "pull_request:46968|pull_request:47019", + "pull_request:46707|pull_request:46968", + "pull_request:46912|pull_request:47025", + "pull_request:46707|pull_request:46957", + "pull_request:46707|pull_request:47019", + "pull_request:46968|pull_request:47062", + "pull_request:46961|pull_request:47019", + "pull_request:46196|pull_request:46965", + "pull_request:46658|pull_request:47062", + "pull_request:46912|pull_request:47078", + "pull_request:46965|pull_request:47062", + "pull_request:43451|pull_request:46995", + "pull_request:46836|pull_request:46995", + "pull_request:46196|pull_request:47025", + "pull_request:46266|pull_request:47034", + "pull_request:46707|pull_request:47025", + "pull_request:45694|pull_request:47069", + "pull_request:43451|pull_request:47079", + "pull_request:46836|pull_request:46965", + "pull_request:46836|pull_request:46961", + "pull_request:46995|pull_request:47003", + "pull_request:47035|pull_request:47069", + "pull_request:46957|pull_request:46961" + ], + "prepared_review_unit_hash": "237bb7d60a04a3dba0e4cf1fdf7b2a43ea68274f8424b347e3d846e92fcc7ec0", + "component_budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 + }, + "budget": { + "node_count": 45, + "item_count": 45, + "soft_pair_count": 72, + "serialized_chars": 126069, + "estimated_input_tokens": 31518, + "estimated_eval_tokens": 63292 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a duplicate set: it mixes unrelated model additions, bug fixes, tests, docs, and automation changes. None of the soft edges look like the same underlying fix or change, so all are rejected.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the items are mostly unrelated fixes, feature additions, and docs/test tweaks rather than one duplicate change.", + "best_issue_reason": null, + "best_pr_reason": "No single PR cleanly represents the cluster as a duplicate group; the highest-scope changes still overlap only loosely with a few neighbors.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46903", + "right": "pull_request:46958", + "accept": false, + "reason": "Same model family, but different loss implementations and different bug paths: shift-label handling vs manual cross-entropy." + }, + { + "left": "pull_request:46950", + "right": "pull_request:46951", + "accept": false, + "reason": "Mamba2 OOM/kernel behavior is unrelated to PEFT checkpoint loading and mismatched-size handling." + }, + { + "left": "pull_request:46950", + "right": "pull_request:47057", + "accept": false, + "reason": "Mamba2 runtime/OOM fix vs DeepSeek V3.2 test offload changes; different models and failures." + }, + { + "left": "pull_request:46927", + "right": "pull_request:47054", + "accept": false, + "reason": "GLM46V input-id dtype handling is unrelated to MiniMax-M3-VL load/offload test setup." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46951", + "accept": false, + "reason": "GLM46V input casting and PEFT loading are different code paths and different problems." + }, + { + "left": "pull_request:46951", + "right": "pull_request:46971", + "accept": false, + "reason": "Checkpoint loading fix vs Mamba2 integration-test dtype/offload workaround; not the same change." + }, + { + "left": "pull_request:46949", + "right": "pull_request:46971", + "accept": false, + "reason": "Whisper cache-state initialization is unrelated to Mamba2 test loading changes." + }, + { + "left": "pull_request:46951", + "right": "pull_request:47057", + "accept": false, + "reason": "PEFT loading behavior vs DeepSeek V3.2 offload test fixes are distinct issues." + }, + { + "left": "pull_request:46941", + "right": "pull_request:46970", + "accept": false, + "reason": "DAC output-mismatch model fixes are unrelated to MusicGen Melody test expectations." + }, + { + "left": "pull_request:47056", + "right": "pull_request:47058", + "accept": false, + "reason": "Whisper cache initialization and Mamba2 OOM test tweaks do not fix the same bug." + }, + { + "left": "pull_request:47006", + "right": "pull_request:47058", + "accept": false, + "reason": "Gemma3 generation-cache logic and Mamba2 OOM test adjustments are different problems." + }, + { + "left": "pull_request:46927", + "right": "pull_request:47055", + "accept": false, + "reason": "GLM46V input casting is unrelated to EdgeTam processor/backbone kwarg cleanup." + }, + { + "left": "pull_request:46908", + "right": "pull_request:47006", + "accept": false, + "reason": "DeepSeek V3.2 offload testing and Gemma3 generation-cache behavior are unrelated." + }, + { + "left": "pull_request:46908", + "right": "pull_request:46928", + "accept": false, + "reason": "DeepSeek V3.2 load/offload tests and BLIP2 tokenizer allowlisting are different changes." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46926", + "accept": false, + "reason": "EdgeTam post-processing kwarg handling is unrelated to MiniMax-M3-VL load/offload testing." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46941", + "accept": false, + "reason": "GLM46V input-id casting and DAC output-mismatch fixes target different models and bugs." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46986", + "accept": false, + "reason": "GLM46V input casting and GLM OCR output mismatch are separate model-specific fixes." + }, + { + "left": "pull_request:46986", + "right": "pull_request:47058", + "accept": false, + "reason": "GLM OCR output-shape mismatch and Mamba2 OOM handling are unrelated." + }, + { + "left": "pull_request:46926", + "right": "pull_request:46928", + "accept": false, + "reason": "MiniMax-M3-VL offload testing and BLIP2 tokenizer registration are unrelated changes." + }, + { + "left": "pull_request:46928", + "right": "pull_request:47057", + "accept": false, + "reason": "BLIP2 tokenizer allowlisting does not address DeepSeek V3.2 loading/offload issues." + }, + { + "left": "pull_request:46926", + "right": "pull_request:46941", + "accept": false, + "reason": "MiniMax-M3-VL test offload and DAC output-mismatch fixes are different code paths." + }, + { + "left": "pull_request:46908", + "right": "pull_request:47058", + "accept": false, + "reason": "DeepSeek V3.2 load/offload tests and Mamba2 OOM fixes are unrelated." + }, + { + "left": "pull_request:46928", + "right": "pull_request:47058", + "accept": false, + "reason": "BLIP2 tokenizer config and Mamba2 OOM test fixes are unrelated." + }, + { + "left": "pull_request:46928", + "right": "pull_request:47054", + "accept": false, + "reason": "BLIP2 tokenizer allowlisting is unrelated to MiniMax-M3-VL load/offload test changes." + }, + { + "left": "pull_request:46941", + "right": "pull_request:46986", + "accept": false, + "reason": "DAC output-mismatch fixes and GLM OCR output-mismatch fixes are different models and bugs." + }, + { + "left": "pull_request:46941", + "right": "pull_request:47006", + "accept": false, + "reason": "DAC tensor-output fixes are unrelated to Gemma3 generation config behavior." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46961", + "accept": false, + "reason": "Pytest v9 fp32-module handling and Serge normalization workflow are unrelated changes." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46968", + "accept": false, + "reason": "Pytest v9 behavior fix is unrelated to docs autolink cleanup." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46974", + "accept": false, + "reason": "Workflow normalization automation and MusicGen generate typo fix are different changes." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46974", + "accept": false, + "reason": "CohereASR loss handling and MusicGen generate typo are unrelated." + }, + { + "left": "pull_request:46954", + "right": "pull_request:46971", + "accept": false, + "reason": "Pytest v9 fp32-module fix is unrelated to Mamba2 test loading/OOM changes." + }, + { + "left": "pull_request:46971", + "right": "pull_request:46974", + "accept": false, + "reason": "Mamba2 OOM test setup and MusicGen generate typo are unrelated." + }, + { + "left": "pull_request:46958", + "right": "pull_request:46968", + "accept": false, + "reason": "CohereASR training-loss fix and docs autolink cleanup are unrelated." + }, + { + "left": "pull_request:46968", + "right": "pull_request:46974", + "accept": false, + "reason": "Docs autolink cleanup and MusicGen generate typo fix are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46968", + "accept": false, + "reason": "Serge workflow automation and docs autolink cleanup are unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47078", + "accept": false, + "reason": "Both are Japanese docs PRs, but one adds DistilBERT docs while the other tweaks wording; not the same change." + }, + { + "left": "pull_request:46266", + "right": "pull_request:46957", + "accept": false, + "reason": "DINO conversion/model modernization and finegrained FP8 generation changes do not address the same bug." + }, + { + "left": "pull_request:47019", + "right": "pull_request:47025", + "accept": false, + "reason": "Masking-utils causal-mask regression fix and Diffusion Gemma test updates are different scopes." + }, + { + "left": "pull_request:47062", + "right": "pull_request:47077", + "accept": false, + "reason": "IMDb link fix in sequence-classification docs is unrelated to Japanese DistilBERT docs translation." + }, + { + "left": "pull_request:47034", + "right": "pull_request:47035", + "accept": false, + "reason": "Neuron training experiment scaffolding and torch-compile backend defaulting are related only by hardware, not the same fix." + }, + { + "left": "pull_request:47019", + "right": "pull_request:47069", + "accept": false, + "reason": "Mask-return contract regression and num_labels returning None are unrelated API changes." + }, + { + "left": "pull_request:46957", + "right": "pull_request:46965", + "accept": false, + "reason": "Finegrained FP8 generation optimization and DeepSeek V4 use_cache support are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47062", + "accept": false, + "reason": "DINO modernization and IMDb docs-link fixes are unrelated." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47025", + "accept": false, + "reason": "RoPE test improvements and Diffusion Gemma test-case changes do not fix the same concrete issue." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47078", + "accept": false, + "reason": "RoPE test changes and Japanese docs wording updates are unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "wav2vec2 SDPA dispatch and flash-linear-attention MLU support are different backend-specific changes." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46995", + "accept": false, + "reason": "Unlimited OCR model addition is unrelated to flash-linear-attention availability checks." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47025", + "accept": false, + "reason": "wav2vec2 SDPA backend support and Diffusion Gemma tests are unrelated." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47034", + "accept": false, + "reason": "DINO modernization and Neuron SFT experiment scripts are unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:47025", + "accept": false, + "reason": "Native FSDP2 migration and Diffusion Gemma test fixes are unrelated." + }, + { + "left": "pull_request:45694", + "right": "pull_request:47069", + "accept": false, + "reason": "Trainer batch-size split_batches fix and num_labels None typing change are different subsystems." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47079", + "accept": false, + "reason": "Molmo2 model addition and AltCLIP text embedding resize test are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46965", + "accept": false, + "reason": "Unlimited OCR model addition and DeepSeek V4 cache support are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:46961", + "accept": false, + "reason": "Unlimited OCR model addition and Serge normalization workflow are unrelated." + }, + { + "left": "pull_request:46995", + "right": "pull_request:47003", + "accept": false, + "reason": "Flash-linear-attention MLU support and OpenVINO exporter work are different backend integrations." + }, + { + "left": "pull_request:47035", + "right": "pull_request:47069", + "accept": false, + "reason": "Neuron compile-backend defaulting and num_labels return-type typing are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:46961", + "accept": false, + "reason": "Finegrained FP8 generation work and Serge branch normalization workflow 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": 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-40546-48", + "nodes": [ + "pull_request:43451", + "pull_request:44517", + "pull_request:45694", + "pull_request:46196", + "pull_request:46266", + "pull_request:46658", + "pull_request:46707", + "pull_request:46836", + "pull_request:46895", + "pull_request:46907", + "pull_request:46908", + "pull_request:46912", + "pull_request:46926", + "pull_request:46927", + "pull_request:46949", + "pull_request:46950", + "pull_request:46951", + "pull_request:46957", + "pull_request:46958", + "pull_request:46961", + "pull_request:46965", + "pull_request:46968", + "pull_request:46970", + "pull_request:46971", + "pull_request:46986", + "pull_request:47003", + "pull_request:47006", + "pull_request:47025", + "pull_request:47034", + "pull_request:47035", + "pull_request:47054", + "pull_request:47055", + "pull_request:47056", + "pull_request:47057", + "pull_request:47058", + "pull_request:47062", + "pull_request:47069", + "pull_request:47077", + "pull_request:47078" + ], + "soft_pairs": [ + "pull_request:46961|pull_request:47035", + "pull_request:43451|pull_request:47025", + "pull_request:46707|pull_request:47077", + "pull_request:46912|pull_request:47077", + "pull_request:43451|pull_request:46266", + "pull_request:46266|pull_request:47003", + "pull_request:46196|pull_request:46961", + "pull_request:46658|pull_request:46965", + "pull_request:45694|pull_request:47035", + "pull_request:43451|pull_request:47078", + "pull_request:46658|pull_request:46961", + "pull_request:46961|pull_request:46965", + "pull_request:47003|pull_request:47078", + "pull_request:43451|pull_request:44517", + "pull_request:45694|pull_request:46707", + "pull_request:46836|pull_request:47078", + "pull_request:44517|pull_request:46836", + "pull_request:46961|pull_request:47069", + "pull_request:44517|pull_request:47003", + "pull_request:43451|pull_request:46957", + "pull_request:43451|pull_request:46961", + "pull_request:46961|pull_request:47003", + "pull_request:46961|pull_request:47034", + "pull_request:46912|pull_request:46961", + "pull_request:46658|pull_request:46895", + "pull_request:44517|pull_request:47077", + "pull_request:43451|pull_request:47069", + "pull_request:43451|pull_request:45694", + "pull_request:43451|pull_request:46895", + "pull_request:45694|pull_request:47003", + "pull_request:46895|pull_request:47003", + "pull_request:46836|pull_request:46895", + "pull_request:46968|pull_request:47003", + "pull_request:46895|pull_request:46958", + "pull_request:46907|pull_request:47055", + "pull_request:46949|pull_request:47056", + "pull_request:46950|pull_request:46971", + "pull_request:46950|pull_request:47058", + "pull_request:46971|pull_request:47058", + "pull_request:46926|pull_request:47054", + "pull_request:46908|pull_request:47057", + "pull_request:46949|pull_request:46950", + "pull_request:47054|pull_request:47057", + "pull_request:46949|pull_request:46951", + "pull_request:46927|pull_request:46949", + "pull_request:46907|pull_request:46927", + "pull_request:46908|pull_request:47054", + "pull_request:46927|pull_request:46971", + "pull_request:46926|pull_request:47057", + "pull_request:46908|pull_request:46926", + "pull_request:46986|pull_request:47006", + "pull_request:47054|pull_request:47055", + "pull_request:46970|pull_request:47006", + "pull_request:46970|pull_request:46986", + "pull_request:47057|pull_request:47058", + "pull_request:46907|pull_request:47054", + "pull_request:47055|pull_request:47056", + "pull_request:46907|pull_request:47056", + "pull_request:47054|pull_request:47056", + "pull_request:47054|pull_request:47058", + "pull_request:47055|pull_request:47057", + "pull_request:46907|pull_request:47006", + "pull_request:46907|pull_request:46970", + "pull_request:46970|pull_request:47055", + "pull_request:46907|pull_request:46986", + "pull_request:46707|pull_request:47034", + "pull_request:46836|pull_request:47062", + "pull_request:46658|pull_request:46836", + "pull_request:43451|pull_request:46836", + "pull_request:46836|pull_request:47003", + "pull_request:43451|pull_request:47077", + "pull_request:46707|pull_request:46836" + ], + "prepared_review_unit_hash": "2e2973c1b08c68a1b395072efa63082304dd1e5651893f271218830aaf4081cb", + "component_budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 + }, + "budget": { + "node_count": 39, + "item_count": 39, + "soft_pair_count": 72, + "serialized_chars": 115600, + "estimated_input_tokens": 28900, + "estimated_eval_tokens": 58056 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated PRs were grouped together; only a few pairs are true duplicates on the same file/code path. The strongest duplicate signals are Edgetam forward-kwarg filtering, Whisper cache initialization, CohereASR loss shifting, and two test-only load_error fixes.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "Pick pull_request:47055 as the canonical PR: it is the clearest and most complete Edgetam kwarg-filtering fix, matches 46907 on the same model files and forward path, and directly addresses the same concrete bug.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:47055 is the best overall representative among the true duplicates because it fixes a concrete runtime code path in Edgetam rather than a broad subsystem or doc/test-only change, and it subsumes the smaller 46907 patch.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46961", + "right": "pull_request:47035", + "accept": false, + "reason": "Unrelated changes: Serge workflow normalization vs torch.compile backend defaulting to neuron. Different code paths and goals." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47025", + "accept": false, + "reason": "Molmo2 model/docs addition is unrelated to DiffusionGemma test-case adjustments. Different models and problems." + }, + { + "left": "pull_request:46707", + "right": "pull_request:47077", + "accept": false, + "reason": "Native FSDP2 module/migration is unrelated to Japanese DistilBERT docs translation. Not the same change." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47077", + "accept": false, + "reason": "RoPE test coverage changes and Japanese docs wording are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46266", + "accept": false, + "reason": "Molmo2 addition and DINO conversion-mapping modernization are different model families and code paths." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47003", + "accept": false, + "reason": "DINO conversion/model changes do not match the OpenVINO exporter feature work." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46961", + "accept": false, + "reason": "wav2vec2 SDPA dispatch support is unrelated to Serge branch normalization workflow plumbing." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46965", + "accept": false, + "reason": "Step 3.7 Flash model conversion/setup work is unrelated to DeepSeek V4 cache handling." + }, + { + "left": "pull_request:45694", + "right": "pull_request:47035", + "accept": false, + "reason": "Batch-size/split-batches behavior in TrainingArguments is unrelated to Neuron compile backend defaults." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47078", + "accept": false, + "reason": "Molmo2 model addition is unrelated to Japanese docs wording cleanup." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46961", + "accept": false, + "reason": "Step 3.7 Flash work and Serge workflow normalization are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:46965", + "accept": false, + "reason": "Workflow normalization and DeepSeek V4 cache behavior are unrelated changes." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47078", + "accept": false, + "reason": "OpenVINO exporter feature work is unrelated to Japanese documentation wording." + }, + { + "left": "pull_request:43451", + "right": "pull_request:44517", + "accept": false, + "reason": "Molmo2 and Qwen3-TTS are different model additions with different files and tasks." + }, + { + "left": "pull_request:45694", + "right": "pull_request:46707", + "accept": false, + "reason": "TrainingArguments split-batches fix is unrelated to native FSDP2 module migration." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47078", + "accept": false, + "reason": "Unlimited OCR model addition is unrelated to Japanese docs wording." + }, + { + "left": "pull_request:44517", + "right": "pull_request:46836", + "accept": false, + "reason": "Qwen3-TTS and Unlimited OCR are separate model additions, not the same underlying change." + }, + { + "left": "pull_request:46961", + "right": "pull_request:47069", + "accept": false, + "reason": "Workflow normalization and num_labels typing behavior are unrelated." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47003", + "accept": false, + "reason": "Qwen3-TTS model addition is unrelated to OpenVINO exporter support." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46957", + "accept": false, + "reason": "Molmo2 addition is unrelated to finegrained FP8 / decode optimization changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46961", + "accept": false, + "reason": "Molmo2 model addition is unrelated to Serge workflow normalization." + }, + { + "left": "pull_request:46961", + "right": "pull_request:47003", + "accept": false, + "reason": "Workflow normalization and OpenVINO exporter support are unrelated." + }, + { + "left": "pull_request:46961", + "right": "pull_request:47034", + "accept": false, + "reason": "Workflow normalization is unrelated to Neuron SFT experiment scripts and training args changes." + }, + { + "left": "pull_request:46912", + "right": "pull_request:46961", + "accept": false, + "reason": "RoPE test changes and workflow normalization are unrelated." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46895", + "accept": false, + "reason": "Step 3.7 Flash work and CohereASR loss-shift fix are unrelated model changes." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47077", + "accept": false, + "reason": "Qwen3-TTS model addition is unrelated to Japanese DistilBERT docs translation." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47069", + "accept": false, + "reason": "Molmo2 addition is unrelated to the num_labels property typing fix." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45694", + "accept": false, + "reason": "Molmo2 addition is unrelated to TrainingArguments split-batches behavior." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46895", + "accept": false, + "reason": "Molmo2 addition is unrelated to CohereASR training-loss fix." + }, + { + "left": "pull_request:45694", + "right": "pull_request:47003", + "accept": false, + "reason": "TrainingArguments split-batches fix is unrelated to OpenVINO exporter support." + }, + { + "left": "pull_request:46895", + "right": "pull_request:47003", + "accept": false, + "reason": "CohereASR loss-shift fix is unrelated to OpenVINO exporter support." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47003", + "accept": false, + "reason": "Unlimited OCR addition and OpenVINO exporter work are different feature areas." + }, + { + "left": "pull_request:46968", + "right": "pull_request:47003", + "accept": false, + "reason": "Docs autolink cleanup is unrelated to OpenVINO exporter implementation." + }, + { + "left": "pull_request:46895", + "right": "pull_request:46958", + "accept": true, + "reason": "Same CohereASR training-loss double-shift fix: both touch the same model forward path and the same files, with the same issue target 46894." + }, + { + "left": "pull_request:46907", + "right": "pull_request:47055", + "accept": true, + "reason": "Same Edgetam bug: both patches remove processor-only kwargs before calling the backbone in the same model files; 47055 is just the broader follow-up of 46907." + }, + { + "left": "pull_request:46949", + "right": "pull_request:47056", + "accept": true, + "reason": "Same Whisper cache-state bug: both change the same forward path in modeling_whisper.py to initialize cache update state consistently." + }, + { + "left": "pull_request:46950", + "right": "pull_request:46971", + "accept": false, + "reason": "Both mention Mamba2 OOM, but one changes model execution and the other only adjusts test loading arguments; not the same concrete fix." + }, + { + "left": "pull_request:46950", + "right": "pull_request:47058", + "accept": false, + "reason": "One PR changes Mamba2 kernel behavior, the other only tweaks integration tests and cleanup; different code paths." + }, + { + "left": "pull_request:46926", + "right": "pull_request:47054", + "accept": true, + "reason": "Same MiniMax-M3-VL load_error workaround: both modify the same integration test to add an offload folder for device_map='auto' loading." + }, + { + "left": "pull_request:46908", + "right": "pull_request:47057", + "accept": true, + "reason": "Same DeepSeekV3.2 load_error workaround: both patch the same test file to use a temporary offload folder for auto device-mapped loading." + }, + { + "left": "pull_request:46949", + "right": "pull_request:46950", + "accept": false, + "reason": "Different models and different fixes: Whisper cache handling versus Mamba2 OOM behavior." + }, + { + "left": "pull_request:46927", + "right": "pull_request:46949", + "accept": false, + "reason": "GLM46V input-id casting and Whisper cache initialization are unrelated model fixes." + }, + { + "left": "pull_request:46908", + "right": "pull_request:47054", + "accept": false, + "reason": "DeepSeekV3.2 and MiniMax-M3-VL are different model test fixes, even though both use tempfile/offload_folder." + }, + { + "left": "pull_request:46926", + "right": "pull_request:47057", + "accept": false, + "reason": "MiniMax-M3-VL and DeepSeekV3.2 are different models and separate test workarounds." + }, + { + "left": "pull_request:46970", + "right": "pull_request:47006", + "accept": false, + "reason": "musicgen_melody test-output mismatch fix and gemma3 generation-config fix are unrelated." + }, + { + "left": "pull_request:46986", + "right": "pull_request:47006", + "accept": false, + "reason": "glm_ocr output mismatch tests and gemma3 generation behavior are unrelated." + }, + { + "left": "pull_request:47054", + "right": "pull_request:47055", + "accept": false, + "reason": "MiniMax-M3-VL test load workaround and Edgetam forward-path fix are different models and problems." + }, + { + "left": "pull_request:46970", + "right": "pull_request:47056", + "accept": false, + "reason": "musicgen_melody and Whisper are separate model-specific fixes." + }, + { + "left": "pull_request:46971", + "right": "pull_request:47058", + "accept": false, + "reason": "Both are Mamba2-related, but one is a test loading tweak and the other is a model/kernel change; not the same underlying bug fix." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47003", + "accept": false, + "reason": "Finegrained FP8 decode optimization work is unrelated to OpenVINO exporter support." + }, + { + "left": "pull_request:47055", + "right": "pull_request:47057", + "accept": false, + "reason": "Edgetam forward-path kwarg filtering and DeepSeekV3.2 test loading work are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:47054", + "accept": false, + "reason": "Edgetam fix and MiniMax-M3-VL test workaround are different models and different code paths." + }, + { + "left": "pull_request:46907", + "right": "pull_request:47056", + "accept": false, + "reason": "Edgetam and Whisper fixes target different model implementations." + }, + { + "left": "pull_request:47054", + "right": "pull_request:47056", + "accept": false, + "reason": "MiniMax-M3-VL and Whisper are unrelated model-specific fixes." + }, + { + "left": "pull_request:47054", + "right": "pull_request:47058", + "accept": false, + "reason": "MiniMax-M3-VL test workaround and Mamba2 OOM fix are different code paths." + }, + { + "left": "pull_request:47055", + "right": "pull_request:47057", + "accept": false, + "reason": "Edgetam forward fix and DeepSeekV3.2 test workaround are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:47006", + "accept": false, + "reason": "Edgetam kwarg filtering and gemma3 generation-config changes are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46970", + "accept": false, + "reason": "Edgetam fix and musicgen_melody output mismatch fix are unrelated model changes." + }, + { + "left": "pull_request:46970", + "right": "pull_request:47055", + "accept": false, + "reason": "musicgen_melody test-output fix and Edgetam forward-path fix are unrelated." + }, + { + "left": "pull_request:46907", + "right": "pull_request:46986", + "accept": false, + "reason": "Edgetam forward-path cleanup and glm_ocr output mismatch tests are unrelated." + }, + { + "left": "pull_request:46707", + "right": "pull_request:47034", + "accept": false, + "reason": "Native FSDP2 migration and Neuron SFT experiment scaffolding are different initiatives." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47062", + "accept": false, + "reason": "Unlimited OCR model addition and IMDb link documentation fix are unrelated." + }, + { + "left": "pull_request:46658", + "right": "pull_request:46836", + "accept": false, + "reason": "Step 3.7 Flash model work and Unlimited OCR addition are unrelated model integrations." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47077", + "accept": false, + "reason": "Molmo2 model addition is unrelated to Japanese DistilBERT docs translation." + }, + { + "left": "pull_request:46707", + "right": "pull_request:46836", + "accept": false, + "reason": "FSDP2 migration and Unlimited OCR model addition are unrelated changes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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-40546-48", + "nodes": [ + "pull_request:40546", + "pull_request:43451", + "pull_request:46196", + "pull_request:46658", + "pull_request:46707", + "pull_request:46912", + "pull_request:47003", + "pull_request:47034", + "pull_request:47035" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:43451", + "pull_request:43451|pull_request:46196", + "pull_request:46196|pull_request:47003", + "pull_request:43451|pull_request:46658", + "pull_request:46658|pull_request:47003", + "pull_request:43451|pull_request:47003", + "pull_request:43451|pull_request:47034", + "pull_request:43451|pull_request:47035", + "pull_request:43451|pull_request:46912", + "pull_request:47003|pull_request:47034", + "pull_request:47003|pull_request:47035", + "pull_request:46912|pull_request:47003", + "pull_request:43451|pull_request:46707" + ], + "prepared_review_unit_hash": "cf508ebbcbf6f0cfd27856fe7da89a0e2d47ba730eec7c482ff01be4fbbfe97c", + "component_budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 157, + "serialized_chars": 176063, + "estimated_input_tokens": 44016, + "estimated_eval_tokens": 88288 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 13, + "serialized_chars": 28300, + "estimated_input_tokens": 7075, + "estimated_eval_tokens": 14406 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated model additions, exporter/runtime changes, training tweaks, and test-only work. I do not see any true duplicate PR pairs among them.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear canonical duplicate PR exists; the cluster does not center on one underlying change. If one had to represent the group, PR 43451 (Add Molmo2) is the most complete self-contained change, but it is not a duplicate target for the others.", + "best_issue_reason": null, + "best_pr_reason": "PR 43451 is the most complete and coherent single change in the set (model code, processors, tests, and docs). It is still not globally suitable as a duplicate representative because the rest of the cluster concerns different subsystems and different problems.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "Both are model-addition PRs with docs/auto-mapping churn, but they implement different models (VibeVoice vs Molmo2) and different code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46196", + "accept": false, + "reason": "Molmo2 model integration and wav2vec2 SDPA dispatch are unrelated changes; overlap in audio/model infrastructure is too broad to be a duplicate." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47003", + "accept": false, + "reason": "One adds SDPA dispatch for audio models; the other adds an OpenVINO exporter. Different goals, subsystems, and merge scopes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46658", + "accept": false, + "reason": "Molmo2 integration and Step-3.7 Flash integration are different model families and different implementations, not one underlying bug/change." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47003", + "accept": false, + "reason": "Step-3.7 Flash implementation and OpenVINO exporter work are unrelated; they do not fix the same concrete code-path problem." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47003", + "accept": false, + "reason": "Molmo2 model addition vs exporter support are distinct changes; shared docs/auto files are incidental." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47034", + "accept": false, + "reason": "Molmo2 addition and the Neuron SFT experiment script address different tasks and code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47035", + "accept": false, + "reason": "Molmo2 model addition is unrelated to the torch-compile backend default change for Neuron." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46912", + "accept": false, + "reason": "Molmo2 integration and RoPE test expansion are different kinds of changes; the overlap is only at a very broad framework level." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47034", + "accept": false, + "reason": "OpenVINO exporter support and Neuron fine-tuning scripts are different runtime features with no shared concrete bug." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47035", + "accept": false, + "reason": "Exporter support and Neuron compile-backend defaulting are separate infrastructure changes, not the same fix." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47003", + "accept": false, + "reason": "RoPE testing improvements do not address the exporter change; these are unrelated maintenance areas." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46707", + "accept": false, + "reason": "Molmo2 model addition and native FSDP2 migration are different large-scale changes with no common underlying issue." + } + ] + }, + "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": "338299dbf3edd54e6e21fdf23242105d61bbac0079210ac643368602585e733a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "False positive cluster: one PR adds TimesFM covariate forecasting, the other adds DINOv3 image classification support. They touch different models, code paths, and user-facing features.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: these are unrelated feature additions in different subsystems, not the same underlying change.", + "best_issue_reason": null, + "best_pr_reason": "No best PR for the cluster; neither PR is a suitable representative of the other because they do not overlap in purpose or implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "TimesFM covariate forecasting and DINOv3 image classification are unrelated changes with no shared bug or code-path 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": 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-41855-13", + "nodes": [ + "pull_request:41855", + "pull_request:42668", + "pull_request:43251", + "pull_request:43254", + "pull_request:43424", + "pull_request:43823", + "pull_request:43995", + "pull_request:44827", + "pull_request:45170", + "pull_request:45213", + "pull_request:45655", + "pull_request:46048", + "pull_request:46267" + ], + "soft_pairs": [ + "pull_request:41855|pull_request: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:46267", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "936ef4bf90712da975974bde757cbcc8592d8d4488e12bcb6fbec39bb3ccd123", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34781, + "estimated_input_tokens": 8696, + "estimated_eval_tokens": 17648 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 34781, + "estimated_input_tokens": 8696, + "estimated_eval_tokens": 17648 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly unrelated PRs that only share broad soft-similarity signals. The only clear duplicate-like pair is the two fixed_cross_entropy PRs (43251 and 43254), which both change the same function in the same file for the same issue.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for the whole cluster; most items are unrelated. If a representative is needed, PR 43251 is the stronger canonical for the only true duplicate pair because it explicitly targets issue 43240 and includes the fuller cross-entropy kwargs support.", + "best_issue_reason": null, + "best_pr_reason": "PR 43251 is the best representative of the cluster\u2019s only genuine duplicate pair: it fixes the same cross-entropy behavior as 43254, but is more complete and explicitly tied to the tracking issue.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Different features: Mistral tokenizer missing methods vs Executorch dynamic-shape export tests. No shared bug or change." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Tokenizer helper additions are unrelated to adding MobileLLM model files." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor auto-loading robustness changes are unrelated to adding MobileLLM." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Cross-entropy kwargs passthrough and a new model implementation are unrelated changes." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Loss utility kwargs support does not match Mistral4 test fixes; different code paths and purposes." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Executorch export tests and Mistral4 fixes are unrelated." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Dynamic-shape export testing is unrelated to MobileLLM model addition." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New MobileLLM model code and config-resolution ordering fix are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "MobileLLM model addition and Falcon output-collection refactor are unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor and model-creation skill/docs are unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon refactoring and Mistral4 test fixes address different models and problems." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 tests and layrnorm->layernorm rename/conversion changes are different issues." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46267", + "accept": false, + "reason": "Mistral4 test fixes and chat parsing updates are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 test fixes and TDT loss kernel integration are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 test fixes and config resolution ordering are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "The shared auto/modeling_auto file is incidental; model registration for Sarvam MoE and Mistral4 tests are unrelated changes." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "Adding MobileLLM and creating a model-addition skill are unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both PRs modify the same fixed_cross_entropy function to pass supported kwargs through to nn.functional.cross_entropy for the same underlying issue 43240; they could plausibly be merged into one patch." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet, and the conservative non-match verdicts are supported by the titles/diffs. The 43251/43254 pair is the only clearly mergeable duplicate-like pair: both touch the same fixed_cross_entropy function for the same issue and the patches are compatible as a single bugfix." + }, + "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": "d1ded7571cd9eeb46327d8a6544cc3ff9c3602bcc35d1ef9d4a11cd474a71259", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs both update GitHub Actions dependencies, but they target different action families and different maintenance goals: Node 24 compatibility vs general/latest-version upgrades. They are not the same underlying change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the cleaner representative because it specifically addresses the concrete Node 24 compatibility migration across workflow actions.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the best representative of this cluster since its scope is more clearly defined and tied to a single compatibility objective; PR 42976 is a separate broader dependency bump.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Both are workflow dependency upgrades, but they change different actions and fix different problems. 42975 updates checkout/cache/upload-artifact for Node 24 compatibility, while 42976 updates docker/build-push-action and other actions to newer versions. They would not plausibly merge into one single PR as the same 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": 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": "c45d3406d4ad45a8cfd39329f01f87f7decda1a50bc9e64a825ca719e2ec9bb3", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13045, + "estimated_input_tokens": 3262, + "estimated_eval_tokens": 6780 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13045, + "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 pull requests are separate model-addition efforts for different architectures (ViT NEPA, Molmo, CTSM, Step3p7) and only overlap on generic docs/auto-registry plumbing. They are not duplicates of the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single PR is a clear canonical duplicate target; the cluster contains distinct model integrations rather than one repeated fix/change.", + "best_issue_reason": null, + "best_pr_reason": "None of these is a better duplicate representative than the others, because each introduces a different model and code path.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model additions (ViT NEPA vs Molmo). Shared registry/docs files are generic scaffolding, not evidence of the same code change." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "ViT NEPA and CTSM are unrelated model implementations; overlap is limited to common auto/docs wiring." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM add different models with distinct configs, processors, and tests; not the same underlying bug or change." + }, + { + "left": "pull_request:45490", + "right": "pull_request:46562", + "accept": false, + "reason": "CTSM and Step3p7 are separate model integrations; similarity is only at the framework/registry level, 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": "137c93bc9e9558d4ceb8f73e9db30e05efb5f5bb6abdb92476ce95a24be5648e", + "component_budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19595, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19595, + "estimated_input_tokens": 4899, + "estimated_eval_tokens": 10054 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This is a heterogeneous set of PRs: most are unrelated test/CI/model fixes across different subsystems, and only the two YOLOS PRs appear to address the same concrete device-map mismatch bug.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for the whole set; the only clear duplicate pair is the YOLOS fix, where 46886 is the better anchor because it explicitly targets issue 46869 and describes the device_map=\"auto\" mismatch clearly.", + "best_issue_reason": null, + "best_pr_reason": "PR 46886 is the strongest representative candidate: focused YOLOS bug fix, explicit issue link, and the same underlying problem as 46922.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Different models and different changes; one is Whisper tokenization/tests, the other is CLIP/ViT test edits." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decode behavior vs FP8 dequant/loading fixes; unrelated code paths." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Both mention Whisper, but the concrete fixes differ: tokenizer timestamp/batch decode vs generation watermarking validation and broader CI/test updates." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46922", + "accept": false, + "reason": "Whisper tokenizer fix and YOLOS device-mismatch fix are unrelated." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46539", + "accept": false, + "reason": "Separate test/CI failures in different model suites; no shared underlying bug." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "CLIP/ViT test edits are unrelated to FP8 dequant/loading changes." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "Different fixes in different areas; only broad test churn overlaps, not the same bug." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46922", + "accept": false, + "reason": "Unrelated model test changes versus YOLOS runtime/device placement fix." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46539", + "accept": false, + "reason": "FP8 dequant/loading work is unrelated to Phimoe integration test CI adjustments." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "Different subsystems and concrete bugs; shared test churn is not enough to merge them." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46922", + "accept": false, + "reason": "FP8 dequantization changes and YOLOS device-map fixes are unrelated." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46539", + "accept": false, + "reason": "Generation/watermarking and Whisper changes do not match the Phimoe CI-memory fix." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46922", + "accept": false, + "reason": "Broad generation/Whisper validation changes are unrelated to YOLOS device placement." + }, + { + "left": "pull_request:46539", + "right": "pull_request:46922", + "accept": false, + "reason": "CI memory-limit/test harness changes do not address the YOLOS runtime bug." + }, + { + "left": "pull_request:46886", + "right": "pull_request:46922", + "accept": true, + "reason": "Both PRs fix the same YOLOS device_map=\"auto\" device mismatch in the same file and could plausibly be merged into one change." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The rejected pairs are appropriately treated as unrelated despite superficial filename/model overlap, and the only accepted pair (46886/46922) is supported by the shared YOLOS file and shared issue target, with changes that look mergeable into one bugfix PR. The canonical/representative wording is cautious enough." + }, + "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": "7ba2a46a9e89bfa4ab730471688e5027d5ac7fe75419ec8a21b9cd2c71aed30a", + "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": "All three pull requests are unrelated, intentionally marked \"don't merge,\" and appear to be test/bot/workflow demonstrations rather than the same underlying fix. None should be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: each PR targets a different file set and a different synthetic/demo change, so they do not represent one underlying fix or change.", + "best_issue_reason": null, + "best_pr_reason": "No best PR: there is no shared bug or merged change to preserve, and all three look like separate test PRs rather than duplicate candidates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:46747", + "accept": false, + "reason": "Different code paths and purposes: one changes setup/versioning and a Vit test file, the other changes a workflow trigger. They are not the same fix." + }, + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Overlap is only superficial (shared test file mention). The PRs touch unrelated workflows/utilities versus packaging/version changes, so they cannot be one merged PR." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46747", + "accept": false, + "reason": "One is a broad workflow/test-diff showcase, the other is a single security-gate workflow trigger change. No concrete shared bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "2cc3f5fda8ccd65260a5455877c423fbaa6ae6bb3cf4677ae22dd6c2c77a35df", + "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": "These are two related but distinct PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They share some documentation and pipeline plumbing, but they target different tasks/models and should not be deduplicated.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR for duplication: the two PRs implement different segmentation features and do not appear to be the same underlying change.", + "best_issue_reason": null, + "best_pr_reason": "PR #43613 is the broader, more central change in this cluster: it adds the visual segmentation pipeline with multiple model integrations and tests, making it the better representative PR if one must be chosen.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Reject: these are different pipeline additions (concept segmentation vs visual segmentation) with different target code paths and model/test coverage; they are related in theme but not the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "6cdac516889031a3b6ca966c9d437163ed5d78ad9ea201429a3dd9beb5c1e755", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs both modify `src/transformers/trainer.py`, but they address different trainer extension points: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom trainers. They are not the same underlying change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR #43651 is the stronger standalone change in the cluster because it fixes a concrete training-step behavior and introduces a focused hook for custom trainers; however, it is still unrelated to #43663 and should not be treated as a duplicate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Different code paths and different problems: #43651 is about gradient accumulation loss scaling in `training_step`, while #43663 is about extracting signature-column logic for dataset column filtering. Shared file alone is not enough to treat them 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": "7474a49f668159de578d1030e6826bd8d809f0e71ad812f418f6012f4a389214", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are in `src/transformers/cli/serve.py` and are about TypeAdapter/annotation-related import safety, but they fix different concrete failure modes and are tied to different issues.", + "confidence": 0.69, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43656 is the broader, more representative fix for the CLI annotation NameError theme and has the stronger discussion/context around the file-wide typing cleanup.", + "best_issue_reason": null, + "best_pr_reason": "PR 43656 is the better standalone candidate because it addresses the import-time annotation evaluation problem across the file, not just one annotated parameter.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Related, but not the same underlying bug/change: 43656 postpones annotation evaluation across `serve.py`, while 43842 specifically makes `TypeAdapter` typing safe when pydantic is absent. They target different issue reports and would not be a clean single 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": 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:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", + "pull_request:44019|pull_request:44722", + "pull_request:44086|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44068|pull_request:44116", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44068|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44059|pull_request:44086", + "pull_request:44098|pull_request:44101", + "pull_request:44059|pull_request:44722", + "pull_request:44018|pull_request:44071", + "pull_request: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:44071|pull_request:44086", + "pull_request:44017|pull_request:44071", + "pull_request:44073|pull_request:44161", + "pull_request:44073|pull_request:44101", + "pull_request:44059|pull_request:44071", + "pull_request:44056|pull_request:44101", + "pull_request:44018|pull_request:44059", + "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:44071|pull_request:44129", + "pull_request:43997|pull_request:44129", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154", + "pull_request:44024|pull_request:44076", + "pull_request:44073|pull_request:44074" + ], + "prepared_review_unit_hash": "7812f97bf46c537dbd306616e8375a11470b98a23fc1561868783d534434c0ed", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 76, + "serialized_chars": 84115, + "estimated_input_tokens": 21029, + "estimated_eval_tokens": 42314 + }, + "budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 72, + "serialized_chars": 76937, + "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 theme of standardized output-tracing refactors across many unrelated model families. I only see clear duplicate PRs inside the GPT-J/codegen line, plus a GPT-Neo revision pair; the cross-model pairs should stay separate.", + "confidence": 0.87, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44722 is the strongest representative of the duplicate GPT-J/codegen line: it is the latest and most complete variant, and it overlaps the same concrete gptj/modeling_gptj.py and codegen/modeling_codegen.py output-tracing changes as the earlier GPT-J PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 44722 is the best overall representative because it appears to supersede the earlier GPT-J/codegen output-tracing refactors rather than being a narrower or older partial version.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model families (FNet/CVT vs GPT-J); same general refactor pattern, but not the same underlying bug/change." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "Unrelated architectures and code paths; both standardize outputs, but they are not duplicates." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Same GPT-J output-tracing refactor on overlapping gptj/modeling_gptj.py changes; 44085 looks like a narrower/earlier variant of the same work." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "Different models (RegNet vs EfficientNet); same refactor theme only, not the same change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet and GPT-J/codegen are unrelated model implementations; no shared code-path bug." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "Different model families and distinct files; these are separate refactors." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa v2 are unrelated code paths; not duplicates." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/codegen refactor vs MGP-STR refactor; same umbrella pattern only, not the same concrete change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/codegen and MPT touch different implementations; not the same bug or patch." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo output-tracing refactor in the same file/test area; 44068 appears to be a later revision of the same change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/codegen vs GPT-Neo; same refactor style but unrelated code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo and GPT-J are different model implementations, so this is not a duplicate." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet output tracing and GPT-J/codegen output tracing are separate model-specific changes." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/codegen are unrelated model families." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are different code paths; same theme only." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo vs ResNet; no shared underlying bug." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are unrelated implementations." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/codegen are separate model refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert/XLM are unrelated changes." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo vs Flaubert/XLM; not duplicates." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-2 and MGP-STR differ in model and code path." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert/XLM are unrelated model-specific output-handling refactors." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-2 and GPT-J/codegen are separate model implementations; not mergeable as one PR." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "SegFormer and LongT5 touch different subsystems and concrete changes." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBert and TextNet are unrelated model implementations." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are different code paths; same refactor pattern only." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44101", + "accept": false, + "reason": "VisualBert vs Flaubert/XLM; no shared concrete bug or patch." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert are distinct model refactors." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44071", + "accept": false, + "reason": "MPNet and MPT are unrelated implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/codegen are different model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44059", + "accept": false, + "reason": "GPT-Neo and GPT-2 are separate model-specific changes." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-2 vs GPT-J/codegen; same tracing theme, not the same bug." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", + "accept": false, + "reason": "Flaubert/XLM and LongT5 are unrelated model implementations." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and SegFormer are different model families." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated code paths." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are distinct implementations; no overlap beyond the broad refactor theme." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert/XLM are unrelated model-specific changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are separate code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are unrelated model implementations." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are different concrete changes." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet and LongT5 are unrelated model refactors." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44129", + "accept": false, + "reason": "MPNet and SpeechT5 are different code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are separate model implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44018", + "accept": false, + "reason": "SqueezeBERT and GPT-Neo are unrelated models; not duplicates." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are different model families." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "FNet/CVT and SegFormer are unrelated changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT and DeBERTa v2 are distinct code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa v2 and SpeechT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa v2 and ViTS are different architectures and changes." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are different model implementations." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBert and TextNet are unrelated code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/codegen output-tracing work in the same files; 44722 looks like the later, broader version of 44066." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both are GPT-J output-tracing refactors on the same code path; they overlap enough to be treated as duplicate PRs." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated models." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "FNet/CVT and RegNet are separate model refactors." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "FNet/CVT and SpeechT5 are unrelated code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative on the cross-model pairs. The accepted duplicate pairs are supported by shared files/targets and appear mergeable as single bugfix PRs, with 44722 a reasonable canonical representative of the GPT-J/codegen line." + }, + "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": "19898176a225f9d5e0f396d7554ae99c8303bc7c7c62ab2044f6fda829297702", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 76, + "serialized_chars": 84115, + "estimated_input_tokens": 21029, + "estimated_eval_tokens": 42314 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 4, + "serialized_chars": 12756, + "estimated_input_tokens": 3189, + "estimated_eval_tokens": 6634 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they target different architectures and code paths, so they are not duplicates of one another.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44129 is the strongest representative of the cluster because it has an explicit issue target and a focused, self-contained SpeechT5 output-capture refactor.", + "best_issue_reason": null, + "best_pr_reason": "PR 44129 is the best standalone PR in this set: it is clearly scoped, tied to an issue, and the diff centers on one concrete model implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "Both are output-tracing refactors, but one is MPNet and the other is VisualBert; they touch different model implementations and are separate changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "These are separate model migrations (MPNet vs TextNet), not the same underlying bug or code change." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are unrelated code paths; similarity is only at the level of a shared refactor pattern." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 modify different model files and different forward logic, 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": 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": "5e554becee13f46801ec8a3122348e554c38c6b810fe9b7ebbdc326fc0a51037", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40179, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40179, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs all follow the same broad \u201coutput tracing refactor\u201d theme, but they are model-specific changes in different code paths and should not be deduplicated as the same underlying change. No soft edge looks like a true duplicate/mergeable pair.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "44003 is the broadest single refactor in the set, touching both `mamba` and `falcon_mamba` and introducing the tracing helper wiring; it is the best representative if a single PR must anchor the cluster.", + "best_issue_reason": null, + "best_pr_reason": "44003 is the strongest representative of this cluster because it contains the most substantial output-tracing implementation changes rather than just signature cleanup in one model file.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations (`timm_backbone` vs `upernet`); same refactor pattern, not the same code-path change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Separate model-specific refactors (`timm_backbone` vs `vision_text_dual_encoder`), not one shared bug or mergeable change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "`timm_backbone` and `univnet` are unrelated code paths; similarity is only at the refactor template level." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Different architectures and files; this is not the same underlying tracing bug/change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "`timm_backbone` and `rwkv` are unrelated models, so this is not a duplicate fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "`mobilenet_v1` and `univnet` change different model-specific forward paths; not the same issue." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different subsystems (`mobilenet_v1` vs `upernet`); no shared concrete bug." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "`mobilenet_v1` and `depth_anything` are separate model refactors, not one mergeable change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "The PRs target different encoder/decoder models and code paths; not duplicates." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "`mobilenet_v1` and `speech_encoder_decoder` are unrelated implementations despite similar wording." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Different models (`vision_text_dual_encoder` vs `univnet`); same refactor style, not same change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "These touch unrelated code paths and model families, so they are not duplicates." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "`vision_text_dual_encoder` and `mamba/falcon_mamba` address different model internals; not mergeable as one fix." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and different forward paths; no shared concrete bug." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "`univnet` and `upernet` are separate model implementations with independent refactors." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "The changes affect unrelated code paths, so this is not the same underlying change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "`univnet` and `depth_anything` are different model-specific tracing refactors, not a duplicate fix." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "`upernet` and `rwkv` are unrelated models; the similarity is only in the refactor pattern." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Different architectures and files; no common concrete bug or code-path change." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "`mamba/falcon_mamba` and `superpoint` are unrelated implementations; not the same underlying issue." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "`mamba/falcon_mamba` vs `codegen` is just shared refactor language, not the same fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "`timm_backbone` and `codegen` are different code paths and cannot be merged as one PR." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families (`univnet` vs `codegen`); no shared concrete bug." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "`upernet` and `codegen` are unrelated model refactors, not duplicates." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "`vision_text_dual_encoder` and `codegen` touch different model internals; not the same change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "`depth_anything` and `vision_encoder_decoder` are separate model-specific refactors with no shared code-path bug." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "These affect different model families and are not the same underlying fix." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "`depth_anything` and `superpoint` are different models; shared theme is too broad for duplicate triage." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "`depth_anything` and `rwkv` change unrelated forward paths; not duplicates." + }, + { + "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:44029", + "accept": false, + "reason": "`vision_encoder_decoder` and `rwkv` do not share the same concrete bug or fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder-decoder stacks; same refactor wording, but not the same change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated code paths." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "`vision_encoder_decoder` and `dpr` are different model families, so this is not a duplicate." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "`speech_encoder_decoder` and `superpoint` are separate implementations; not mergeable as one PR." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Different models and different tracing code paths; not the same bug." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "`speech_encoder_decoder` and `dpr` are unrelated model changes." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "`superpoint` and `rwkv` are different model internals; similarity is only superficial." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "These PRs touch unrelated models and should not be merged as duplicates." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "`rwkv` and `dpr` are separate model-specific refactors, not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "2bfc3e14e2fdd17df8894c45b3a7e98739c2872e1c05242c2270dbe1ab95c967", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6552, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6552, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate model-implementation PRs that only overlap on generic Transformers scaffolding (docs, auto-registry, package wiring). They do not appear to fix the same underlying bug or add the same model.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #44662 is the better cluster representative: it is a fuller, non-draft implementation with substantial review activity and a broader set of model, processor, and test files.", + "best_issue_reason": null, + "best_pr_reason": "PR #44662 is the strongest standalone PR to keep as reference because it is more complete and mature than the draft Voxtral TTS PR.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different model additions with different code paths: PenguinVL vs Voxtral TTS. The shared auto/docs files are generic integration points, not evidence of the same change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "d14c82a8ca98e3980308e8538c247e04166c8c0fc173004c6744999b3dbf71e7", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are related around module/kernel fusion, but they are not the same change: #44979 adds the core Module Fusion API and tests, while #45363 builds a specific n-to-1 kernel-fusion path on top of KernelConfig with integrations and an example.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "#44979 is the more fundamental, general-purpose API change: it introduces the module fusion abstraction itself and accompanying tests, making it the better canonical representative.", + "best_issue_reason": null, + "best_pr_reason": "#44979 is the best representative of the cluster because it is the core infrastructure PR; #45363 is a narrower feature built on top of that abstraction.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Related but distinct: one adds the fusion API primitives, the other implements kernel-config-driven n-to-1 fusion integration. They do not look like the same underlying code change and 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": 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": "bc710f9be9e0e4a34a366571b66b7a64b5beaff3c156e7867c39907266d151ec", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests address the same AttributeError in `_patch_mistral_regex` by switching from `tokenizer.backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer` in the same code path. PR 45317 is the fuller version because it adds tests as well.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same fix as 45086 plus test coverage, making it the more complete change set.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the strongest representative: same bug fix, same file, same issue target, and additional validation via tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same concrete bug fix in the same function and file; 45317 is essentially 45086 plus tests, so they are duplicative and could be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "40d3ff98d4cf267c16ccf2cb4a86d4ba842ea256f94421900a9518766237fac2", + "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": "These PRs are related to Nemotron H model support, but they target different levels of the change: one is a narrower refactor/conversion-mapping update, the other is a broader model split into dense/sparse variants with docs and tests. They should not be clustered as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:45569 is the broader, more complete implementation of the Nemotron H redesign, covering the dense/sparse model split plus docs/tests; pull_request:45115 is a narrower inheritance/conversion refactor.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:45569 is the best representative PR for this cluster because it contains the larger user-facing architectural change and supporting documentation/test updates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Related Nemotron H work, but not the same concrete change: 45115 focuses on conversion/inheritance refactoring, while 45569 introduces separate dense/sparse model packages and docs/tests. They would not plausibly merge into one duplicate PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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": "3cde588e9554b04aa342e3f43da60395aa2cd3f650d7d3dced0e8ce694aa857d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6090, + "estimated_input_tokens": 1523, + "estimated_eval_tokens": 3302 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6090, + "estimated_input_tokens": 1523, + "estimated_eval_tokens": 3302 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are closely related but not duplicates: #45332 is the broader end-to-end heterogeneous model support, while #45333 is a narrower config-only subset. They share docs and configuration plumbing, but their scopes diverge materially in modeling/runtime changes.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45332 is the more representative PR for the cluster because it covers the full heterogeneous model feature set, including modeling, masking, cache utilities, integrations, docs, and tests, whereas #45333 only covers per-layer configuration support.", + "best_issue_reason": null, + "best_pr_reason": "#45332 is the best global PR to keep as the canonical representative because it implements the broader, more complete change set; #45333 is a partial follow-on focused on configuration only.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "They overlap on heterogeneous configuration/docs, but #45332 also changes modeling, masking, cache, and integration code paths. #45333 is config-only, so these are related changes rather than 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": 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": "07c765dcaf16f0a1eba7a72926de65ab4447c1761d1d1eed6d6bb85dde8eda6f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related typing-related PRs, but they are not the same underlying change: one is a broad type-checking sweep across many transformers modules, while the other adds targeted typing helper protocols and compatibility shims.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45415 is the broader, more representative typing change: it updates type-checking config and touches many runtime modules in one coherent sweep.", + "best_issue_reason": null, + "best_pr_reason": "PR 45415 is the best single representative of the cluster because its scope and intent are clearer and more central than the narrower helper-typing update in 45425.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Do not merge: 45415 is a wide type-checking pass across many files, while 45425 adds specific protocol/type-alias helpers in _typing.py and related integrations. They share a typing theme and a few files, but not the same concrete bug/change or mergeable PR scope." + } + ] + }, + "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": "f59795662437af70fa5efd159cc666810c793231efad818b41fc6c27193bbb85", + "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": "Both PRs are about expert-parallel / distributed loading behavior, but they address different failure modes: one fixes serialization and ZeRO-3 loading metadata, the other fixes shard wrapping to prevent expert weights from being overwritten under FSDP2/broadcast. They are related in subsystem, not the same underlying bug or change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear duplicate canonical PR: these PRs fix different distributed-loading bugs and should remain separate.", + "best_issue_reason": null, + "best_pr_reason": "PR #45662 is the stronger representative of the cluster because it targets a concrete runtime corruption bug in the loading path and has broader code-path coverage, though it is not a duplicate of #45548.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Reject: both involve EP/distributed loading, but #45548 fixes config serialization and ZeRO-3/accelerate launch loading, while #45662 fixes post-shard wrapping to stop experts being overwritten under FSDP2. Different concrete bugs and patches." + } + ] + }, + "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": "bd9d81806ac3fc92dce5d027a992ac788d4dd4790e7bab98330b4d1c5b856f06", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4672, + "estimated_input_tokens": 1168, + "estimated_eval_tokens": 2592 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are about making captured debug output from patched tests safe under pytest-xdist and easier to collect in CI, but they are not the same concrete change: one focuses on the logging helper, while the other also changes the workflow that aggregates and prints the artifacts.", + "confidence": 0.78, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the broader end-to-end fix: xdist-safe artifact writing plus CI-side collection/printing of multiple captured_info files.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the best representative of the cluster since it includes the core helper change and the downstream workflow handling needed to preserve the debug logs in CI.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": false, + "reason": "Related, but not the same PR change: 45639 only hardens the testing helper\u2019s file output, while 45645 additionally changes the CI workflow to gather and print multiple captured_info files. That extra workflow behavior makes them different concrete fixes rather than a single mergeable 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": 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-45932-2", + "nodes": [ + "pull_request:45932", + "pull_request:46100" + ], + "soft_pairs": [ + "pull_request:45932|pull_request:46100" + ], + "prepared_review_unit_hash": "746c0332085adb1e52ce51a6b661afde007397b60fd161fba02feee1c8aebf1e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two PRs are not duplicates: one mixes CI/telemetry changes with model config adjustments, while the other is a test-migration PR adding DDP/kernel test infrastructure. They share only broad \"do not merge\"/testing vibes, not the same underlying code-path change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "Neither PR is a good canonical duplicate target; they address different concerns and should remain separate. If forced, 46100 is the more self-contained change, but it is not a duplicate of 45932.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "Different changesets and objectives: 45932 touches CI secrets plus model embedding/config code, while 46100 is a test migration adding DDP config and new kernel/ViT tests. No shared concrete bug or mergeable combined fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 25, + "component_count": 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": "8754d528c860337f1224d840dcec7e19c2e253452f71863df40305a3c6eedd89", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5525, + "estimated_input_tokens": 1382, + "estimated_eval_tokens": 3020 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5525, + "estimated_input_tokens": 1382, + "estimated_eval_tokens": 3020 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The pair is not a true duplicate: PR #46126 is a focused EP/trainer integration change, while PR #46269 is a broad distributed branch base with many unrelated distributed/FSDP/CI edits. The shared filenames are incidental overlap, not the same underlying fix or feature.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR #46126 is the more specific, coherent change centered on EP routing and Trainer integration; PR #46269 is too broad and mixed to serve as a canonical duplicate representative.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Different scope and purpose: #46126 adds router-parallel/Trainer integration, while #46269 is a large distributed base branch spanning FSDP, CI, and many unrelated files. They do not look like the same concrete code-path 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": 26, + "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": "22e3bbdcca07d1baaafd975ba2f030febcb738a8c418209799ad9596cf8bfb51", + "component_budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26679, + "estimated_input_tokens": 6670, + "estimated_eval_tokens": 13596 + }, + "budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26679, + "estimated_input_tokens": 6670, + "estimated_eval_tokens": 13596 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All items are small, unrelated workflow/security chore PRs. The only dense similarity cluster is the pr-ci-security-gate.yml group, but those PRs make different changes and should not be merged as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true duplicate PR stands out; 46554 is the most substantive and central workflow-security change, so it is the best representative if one must be chosen.", + "best_issue_reason": null, + "best_pr_reason": "46554 has the broadest security-gate logic change and is the most representative anchor for the workflow-security subgroup, but it is still distinct from the nearby PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46138", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflow files and different changes: stdin piping for self-comment CI vs PR security-gate logic." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflows and different code paths; no shared underlying bug/change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46563", + "accept": false, + "reason": "Self-comment CI message handling is unrelated to security-gate workflow notes." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46716", + "accept": false, + "reason": "Unrelated workflow updates in different files; not the same fix." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflows: message input handling vs build-ci-docker image tag detection." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "SSH runner hardening is unrelated to benchmark_v2 dispatch logic." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "Different workflows and concerns; SSH runner vs anti-slop permissions." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "SSH runner changes do not match assign-reviewers checkout behavior." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflows and different security concerns; not one concrete bug." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46558", + "accept": false, + "reason": "SSH runner env cleanup is unrelated to PR CI security-gate permissions/approval logic." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Different workflow jobs with different behaviors; only broad CI similarity." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Anti-slop permissions and assign-reviewers checkout are distinct workflow changes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflows and fixes; anti-slop vs build image tagging." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46554", + "accept": false, + "reason": "Anti-slop workflow permissions change is unrelated to PR CI security gate changes." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Assign-reviewers base-ref checkout and benchmark dispatch are unrelated changes." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflow files and different code paths." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46554", + "accept": false, + "reason": "Assign-reviewers checkout fix is unrelated to PR CI security-gate logic." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark dispatch support and build image tagging are different changes." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46554", + "accept": false, + "reason": "Benchmark_v2 trigger update is unrelated to PR CI security gate behavior." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46554", + "accept": false, + "reason": "Build-ci-docker input quoting and PR security-gate changes are distinct." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflow purpose: build image tag selection vs security-gate permissions." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46558", + "accept": false, + "reason": "Same file, but different security-gate edits affecting different parts of the workflow." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46563", + "accept": false, + "reason": "Different changes in the same workflow file; not the same concrete fix." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46716", + "accept": false, + "reason": "Both touch the security-gate file, but they change different trust/permission settings." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46563", + "accept": false, + "reason": "Same workflow file, but permissions/approval logic vs review-comment notes are distinct." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46716", + "accept": false, + "reason": "Distinct edits to the same workflow; not a duplicate change." + }, + { + "left": "pull_request:46563", + "right": "pull_request:46716", + "accept": false, + "reason": "Commentary about reusable workflow safety vs secrets handling are different changes." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46563", + "accept": false, + "reason": "Build-ci-docker tagging logic is unrelated to security-gate documentation/comments." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46563", + "accept": false, + "reason": "SSH runner hardening is unrelated to security-gate comments." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46716", + "accept": false, + "reason": "Build image tag logic and PR CI secrets handling are unrelated." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46716", + "accept": false, + "reason": "SSH runner setup changes do not match PR CI secrets handling." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 27, + "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": 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-46281-2", + "nodes": [ + "pull_request:46281", + "pull_request:46890" + ], + "soft_pairs": [ + "pull_request:46281|pull_request:46890" + ], + "prepared_review_unit_hash": "18b0289cc6228dd3645d0530cd09b318137905973fa4f6523890424071daea2f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4279, + "estimated_input_tokens": 1070, + "estimated_eval_tokens": 2396 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4279, + "estimated_input_tokens": 1070, + "estimated_eval_tokens": 2396 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate security hardening PRs that both address path traversal, but in different checkpoint-loading paths and with different inputs. They are related thematically, not duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: 46281 hardens sharded checkpoint filename resolution in `hub.py`, while 46890 validates `transformers_weights` in `modeling_utils.py`; they target different code paths and different attacker-controlled fields.", + "best_issue_reason": null, + "best_pr_reason": "No single best PR for deduping; both are legitimate but distinct fixes, so neither should represent the other.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46281", + "right": "pull_request:46890", + "accept": false, + "reason": "Different underlying bugs and code paths: one validates shard files from a weight map in `src/transformers/utils/hub.py`, the other blocks traversal via `transformers_weights` in `src/transformers/modeling_utils.py`. They would not plausibly merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 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-46574-2", + "nodes": [ + "pull_request:46574", + "pull_request:46580" + ], + "soft_pairs": [ + "pull_request:46574|pull_request:46580" + ], + "prepared_review_unit_hash": "385578fbe1919fad756b03f5af5c3d712d2f18c6e5183534cc4f60a73f6caf21", + "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 unrelated pull requests: one fixes CodeLlama decode-time space stripping, the other fixes DeepSeek checkpoints loading under the wrong tokenizer class.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "pull_request:46574 is the stronger standalone bug fix: it targets one concrete tokenizer decode path and includes a direct issue link and test coverage. pull_request:46580 is also valid, but it is a separate AutoTokenizer routing fix for different checkpoints.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46574", + "right": "pull_request:46580", + "accept": false, + "reason": "Different underlying problems and code paths: CodeLlama decode stripping vs DeepSeek checkpoint-to-tokenizer-class selection. They are 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": 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-46781-3", + "nodes": [ + "pull_request:46781", + "pull_request:46993", + "pull_request:46994" + ], + "soft_pairs": [ + "pull_request:46781|pull_request:46993", + "pull_request:46993|pull_request:46994" + ], + "prepared_review_unit_hash": "4410e1abf58e9736861a2fce68a41e1265d364f7ddabfe4fdf9f3a094ed3f076", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 6604, + "estimated_input_tokens": 1651, + "estimated_eval_tokens": 3558 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 6604, + "estimated_input_tokens": 1651, + "estimated_eval_tokens": 3558 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three pull requests are separate fixes in adjacent generation/cache validation areas, but they address different code paths and failure modes, so they should not be clustered as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR #46994 is the clearest standalone fix in the set: it adds explicit validation for empty entries in `bad_words_ids` / `sequence_bias` and tightens the corresponding error message. The other PRs solve unrelated bugs in prompt lookup decoding and QuantizedCache layer support.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46781", + "right": "pull_request:46993", + "accept": false, + "reason": "Different subsystems and bugs: one is early rejection of unsupported QuantizedCache layer types, the other is prompt lookup decoding exceeding `max_length`. They do not target the same underlying behavior." + }, + { + "left": "pull_request:46993", + "right": "pull_request:46994", + "accept": false, + "reason": "Both are generation-related, but they fix different problems: prompt lookup candidate length calculation vs validation of empty token lists in logits processors. 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": 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-47016-3", + "nodes": [ + "pull_request:47016", + "pull_request:47048", + "pull_request:47066" + ], + "soft_pairs": [ + "pull_request:47016|pull_request:47048", + "pull_request:47016|pull_request:47066", + "pull_request:47048|pull_request:47066" + ], + "prepared_review_unit_hash": "a4232945901083c549cd2fed0926b1e54b41b28068d7476c3f08919efec65eae", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7337, + "estimated_input_tokens": 1835, + "estimated_eval_tokens": 3926 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7337, + "estimated_input_tokens": 1835, + "estimated_eval_tokens": 3926 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three PRs address the same class of failure: Hungarian matching can crash when the cost matrix contains NaN/Inf, and each PR adds a nan_to_num guard in model-specific object-detection loss code. PR 47048 is the best representative because it is the broadest of the set and most centrally covers the issue pattern.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47048 is the strongest canonical representative: it fixes the exact crash mode with a minimal guard in the shared DETR-style matcher path and covers two model families, making it a good anchor for the duplicate cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR 47048 is the best single PR to represent the cluster because it is broad, directly targeted at the non-finite cost-matrix crash, and has the clearest generalizable fix pattern among the three.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47016", + "right": "pull_request:47048", + "accept": false, + "reason": "Same symptom, but different model-specific code paths (RT-DETR vs DETR/Grounding DINO). This looks like parallel hardening of related matchers rather than the same concrete change that should be merged as one PR." + }, + { + "left": "pull_request:47016", + "right": "pull_request:47066", + "accept": false, + "reason": "Both fix Hungarian matcher infeasibility, but they touch different model families and different loss implementations. That is too broad to treat as one duplicate PR pair under a strict code-path standard." + }, + { + "left": "pull_request:47048", + "right": "pull_request:47066", + "accept": true, + "reason": "These two PRs make the same concrete fix pattern in the same kind of Hungarian-matching loss code: guard non-finite costs before scipy assignment. They are close enough in mechanism and scope to plausibly be merged into one broader 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": 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-47026-2", + "nodes": [ + "pull_request:47026", + "pull_request:47046" + ], + "soft_pairs": [ + "pull_request:47026|pull_request:47046" + ], + "prepared_review_unit_hash": "fca1e7916fb4c3b26c55bcc162ae2ee4d2d8d21ae5949053462667a5c29a16e9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4442, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4442, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests appear to address the same xLSTM train-mode crash around `return_last_states=False` in `modeling_xlstm.py`; the second PR looks like a broader, more complete fix with tests and an additional `use_cache` mention.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47046 is the better canonical PR: it targets the same xLSTM forward-path crash, adds tests, and appears to subsume the narrower fix in 47026.", + "best_issue_reason": null, + "best_pr_reason": "PR 47046 is the strongest representative because it covers the same bug and looks more complete/validated than 47026.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47026", + "right": "pull_request:47046", + "accept": true, + "reason": "Same underlying xLSTM train-mode crash path in the same file, with the same issue target; 47046 looks like the more complete version of the 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-47032-6", + "nodes": [ + "pull_request:47032", + "pull_request:47045", + "pull_request:47051", + "pull_request:47060", + "pull_request:47074", + "pull_request:47076" + ], + "soft_pairs": [ + "pull_request:47032|pull_request:47051", + "pull_request:47032|pull_request:47045", + "pull_request:47045|pull_request:47076", + "pull_request:47045|pull_request:47074", + "pull_request:47051|pull_request:47074", + "pull_request:47051|pull_request:47060", + "pull_request:47060|pull_request:47074", + "pull_request:47060|pull_request:47076", + "pull_request:47074|pull_request:47076" + ], + "prepared_review_unit_hash": "c19d4d7da4c86054340120ffba43860549fbeeec21a3fa506d13c3fc33215a9b", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 9, + "serialized_chars": 15764, + "estimated_input_tokens": 3941, + "estimated_eval_tokens": 8138 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 9, + "serialized_chars": 15764, + "estimated_input_tokens": 3941, + "estimated_eval_tokens": 8138 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are thematically related to multimodal pipeline/processor fixes, but they address distinct code paths: stop-sequence handling, generation kwarg forwarding, model-specific post-processing, input key normalization, audio format validation, and Gemma4 placeholder validation. I did not find any true duplicate pairs.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47051 is the broadest representative of the cluster: it touches the shared multimodal pipeline argument-routing path and spans both AnyToAny and ImageTextToText, making it the most central PR to anchor this group.", + "best_issue_reason": null, + "best_pr_reason": "PR 47051 is the best single representative because it targets the common multimodal pipeline layer rather than a single model-specific processor, and its scope is the most central among the set.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47032", + "right": "pull_request:47045", + "accept": false, + "reason": "Different bugs in different components: image-text-to-text stop-sequence handling vs Voxtral audio-array format handling." + }, + { + "left": "pull_request:47032", + "right": "pull_request:47051", + "accept": false, + "reason": "Related only at a high level; 47032 fixes explicit stop_sequence conversion, while 47051 is about routing direct generation kwargs through multimodal pipelines." + }, + { + "left": "pull_request:47045", + "right": "pull_request:47076", + "accept": false, + "reason": "Both are processor changes, but they fix unrelated validation paths: Voxtral audio encoding vs Gemma4 audio/text placeholder construction." + }, + { + "left": "pull_request:47045", + "right": "pull_request:47074", + "accept": false, + "reason": "Distinct multimodal input problems in different code paths; Voxtral audio format encoding is unrelated to AnyToAny key normalization." + }, + { + "left": "pull_request:47051", + "right": "pull_request:47074", + "accept": false, + "reason": "Shared pipeline file but different underlying changes: kwarg forwarding/generation config handling vs input key normalization for image/video aliases." + }, + { + "left": "pull_request:47051", + "right": "pull_request:47060", + "accept": false, + "reason": "One is generic pipeline generation-argument plumbing; the other is Qwen Omni batched post-processing decode behavior." + }, + { + "left": "pull_request:47060", + "right": "pull_request:47074", + "accept": false, + "reason": "Different models and different failure modes: Qwen Omni text post-processing vs AnyToAny multimodal input normalization." + }, + { + "left": "pull_request:47060", + "right": "pull_request:47076", + "accept": false, + "reason": "Separate model-specific processor bugs with no shared concrete code path." + }, + { + "left": "pull_request:47074", + "right": "pull_request:47076", + "accept": false, + "reason": "Both are multimodal preprocessing fixes, but they address different inputs and validations in unrelated processors." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/manifest.json b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..47556bf8d4df441ca958e675c363f48754d9b13e --- /dev/null +++ b/snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260705t060758z", + "artifacts": { + "hybrid": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-07-05T06:18:11Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260705T060758Z", + "variant": "hybrid" +} diff --git a/snapshots/20260705T060758Z/manifest.json b/snapshots/20260705T060758Z/manifest.json index 49b1e05589276a0c3e747441467ce0ae31f5f636..a54874f430012baf048fca427bdc479f9689bc82 100644 --- a/snapshots/20260705T060758Z/manifest.json +++ b/snapshots/20260705T060758Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 0 }, "extracted_at": "2026-07-05T06:07:58Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260705t060758z", + "runs": { + "hybrid-model-20260705t060758z": { + "analysis_id": "hybrid-model-20260705t060758z", + "artifacts": { + "hybrid": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260705T060758Z/analysis-runs/hybrid-model-20260705t060758z/manifest.json", + "model": null, + "published_at": "2026-07-05T06:18:11Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260705T060758Z", "watermark": {